3. Business – Page 604

Dynamic brain-heart interaction in sleep characterized by variational phase-amplitude coupling framework
Engelen, T., Solcà, M. & Tallon-Baudry, C. Interoceptive rhythms in the brain. Nat. Neurosci. 26, 1670–1684 (2023).

PubMed

Google Scholar
Critchley, H. D. & Harrison, N. A. Visceral influences on brain and behavior. Neuron 77, 624–638 (2013).

PubMed

Google Scholar
Jin, H., Li, M., Jeong, E., Castro-Martinez, F. & Zuker, C. S. A body–brain circuit that regulates body inflammatory responses. Nature 630, 695–703 (2024).
Candia-Rivera, D., Catrambone, V., Thayer, J. F., Gentili, C. & Valenza, G. Cardiac sympathetic-vagal activity initiates a functional brain–body response to emotional arousal. Proc. Natl. Acad. Sci. USA 119, e2119599119 (2022).

PubMed
PubMed Central

Google Scholar
Mather, M. & Thayer, J. F. How heart rate variability affects emotion regulation brain networks. Curr. Opin. Behav. Sci. 19, 98–104 (2018).

PubMed
PubMed Central

Google Scholar
Margolis, K. G., Cryan, J. F. & Mayer, E. A. The microbiota-gut-brain axis: from motility to mood. Gastroenterology 160, 1486–1501 (2021).

PubMed

Google Scholar
Silvani, A., Calandra-Buonaura, G., Dampney, R. A. & Cortelli, P. Brain–heart interactions: physiology and clinical implications. Philos. Trans. R. Soc. A 374, 20150181 (2016).

Google Scholar
Varga, S. & Heck, D. H. Rhythms of the body, rhythms of the brain: Respiration, neural oscillations, and embodied cognition. Conscious. Cogn. 56, 77–90 (2017).

PubMed

Google Scholar
Faes, L., Nollo, G., Jurysta, F. & Marinazzo, D. Information dynamics of brain–heart physiological networks during sleep. N. J. Phys. 16, 105005 (2014).

Google Scholar
Catrambone, V., Talebi, A., Barbieri, R. & Valenza, G. Time-resolved brain-to-heart probabilistic information transfer estimation using inhomogeneous point-process models. IEEE Trans. Biomed. Eng. 68, 3366–3374 (2021).

PubMed

Google Scholar
Park, H.-D. & Blanke, O. Heartbeat-evoked cortical responses: Underlying mechanisms, functional roles, and methodological considerations. Neuroimage 197, 502–511 (2019).

PubMed

Google Scholar
Bashan, A., Bartsch, R. P., Kantelhardt, J. W., Havlin, S. & Ivanov, P. C. Network physiology reveals relations between network topology and physiological function. Nat. Commun. 3, 9 (2012).

Google Scholar
de Zambotti, M., Trinder, J., Silvani, A., Colrain, I. M. & Baker, F. C. Dynamic coupling between the central and autonomic nervous systems during sleep: a review. Neurosci. Biobehav. Rev. 90, 84–103 (2018).

PubMed
PubMed Central

Google Scholar
Yao, Y. et al. Cardiovascular baroreflex circuit moonlights in sleep control. Neuron 110, 3986–3999. e3986 (2022).

PubMed

Google Scholar
Narkiewicz, K. & Somers, V. Sympathetic nerve activity in obstructive sleep apnoea. Acta Physiol. Scand. 177, 385–390 (2003).

PubMed

Google Scholar
Kanki, M. et al. Poor sleep and shift work associate with increased blood pressure and inflammation in UK Biobank participants. Nat. Commun. 14, 7096 (2023).

PubMed
PubMed Central

Google Scholar
Koch, E. Die irradiation der pressoreceptorischen kreislaufreflexe. Klin. Woche. 11, 225–227 (1932).

Google Scholar
Bonvallet, M., Dell, P. & Hiebel, G. Tonus sympathique et activité électrique corticale. Electroencephalogr. Clin. Neurophysiol. 6, 119–144 (1954).

PubMed

Google Scholar
Suarez-Roca, H., Mamoun, N., Sigurdson, M. I. & Maixner, W. Baroreceptor modulation of the cardiovascular system, pain, consciousness, and cognition. Compr. Physiol. 11, 1373 (2021).

PubMed
PubMed Central

Google Scholar
Laborde, S., Mosley, E. & Mertgen, A. A unifying conceptual framework of factors associated to cardiac vagal control. Heliyon 4, e01002 (2018).
Rosenbaum, M. & Race, D. Frequency-response characteristics of vascular resistance vessels. Am. J. Physiol. 215, 1397–1402 (1968).

PubMed

Google Scholar
Deegan, P. & McNicholas, W. Pathophysiology of obstructive sleep apnoea. Eur. Resp. J. 8, 1161–1178 (1995).

Google Scholar
Eckberg, D. L. Topical review: the human respiratory gate. J. Physiol. 548, 339–352 (2003).

PubMed
PubMed Central

Google Scholar
Ako, M. et al. Correlation between electroencephalography and heart rate variability during sleep. Psychiatry Clin. Neurosci. 57, 59–65 (2003).

PubMed

Google Scholar
Jurysta, F. et al. A study of the dynamic interactions between sleep EEG and heart rate variability in healthy young men. Clin. Neurophysiol. 114, 2146–2155 (2003).

PubMed

Google Scholar
Faes, L., Kugiumtzis, D., Nollo, G., Jurysta, F. & Marinazzo, D. Estimating the decomposition of predictive information in multivariate systems. Phys. Rev. E 91, 032904 (2015).

Google Scholar
Orjuela-Cañón, A. D., Cerquera, A., Freund, J. A., Juliá-Serdá, G. & Ravelo-García, A. G. Sleep apnea: tracking effects of a first session of CPAP therapy by means of Granger causality. Comput. Methods Prog. Biomed. 187, 105235 (2020).

Google Scholar
Valenza, G. et al. Combining electroencephalographic activity and instantaneous heart rate for assessing brain–heart dynamics during visual emotional elicitation in healthy subjects. Philos. Trans. R. Soc. A 374, 20150176 (2016).

Google Scholar
Catrambone, V., Greco, A., Scilingo, E. P. & Valenza, G. Functional linear and nonlinear brain–heart interplay during emotional video elicitation: A maximum information coefficient study. Entropy 21, 892 (2019).

PubMed Central

Google Scholar
Schulz, S., Haueisen, J., Bär, K.-J. & Voss, A. Altered causal coupling pathways within the central-autonomic-network in patients suffering from schizophrenia. Entropy 21, 733 (2019).

PubMed
PubMed Central

Google Scholar
Schiecke, K. et al. Brain–heart interactions considering complex physiological data: processing schemes for time-variant, frequency-dependent, topographical and statistical examination of directed interactions by convergent cross mapping. Physiol. Meas. 40, 114001 (2019).

PubMed

Google Scholar
Yeh, C.-H., Lo, M.-T. & Hu, K. Spurious cross-frequency amplitude–amplitude coupling in nonstationary, nonlinear signals. Phys. A 454, 143–150 (2016).

Google Scholar
Pittman-Polletta, B., Hsieh, W.-H., Kaur, S., Lo, M.-T. & Hu, K. Detecting phase-amplitude coupling with high frequency resolution using adaptive decompositions. J. Neurosci. Methods 226, 15–32 (2014).

PubMed
PubMed Central

Google Scholar
Catrambone, V. & Valenza, G. Microstates of the cortical brain-heart axis. Hum. Brain Mapp. 44, 5846–5857 (2023).

PubMed
PubMed Central

Google Scholar
Dragomiretskiy, K. & Zosso, D. Variational mode decomposition. IEEE Trans. Signal Process. 62, 531–544 (2013).

Google Scholar
Li, H., Kwong, S., Yang, L., Huang, D. & Xiao, D. Hilbert-Huang transform for analysis of heart rate variability in cardiac health. IEEE-ACM Trans. Comput. Biol. Bioinform. 8, 1557–1567 (2011).

PubMed

Google Scholar
Orini, M., Bailón, R., Mainardi, L. T., Laguna, P. & Flandrin, P. Characterization of dynamic interactions between cardiovascular signals by time-frequency coherence. IEEE Trans. Biomed. Eng. 59, 663–673 (2011).

PubMed

Google Scholar
Candia-Rivera, D., de Vico Fallani, F. & Chavez, M. Robust and time-resolved estimation of cardiac sympathetic and parasympathetic indices. R. Soc. Open Sci. 12, 240750 (2025).

PubMed
PubMed Central

Google Scholar
Tort, A. B., Komorowski, R., Eichenbaum, H. & Kopell, N. Measuring phase-amplitude coupling between neuronal oscillations of different frequencies. J. Neurophysiol. 104, 1195–1210 (2010).

PubMed
PubMed Central

Google Scholar
Aru, J. et al. Untangling cross-frequency coupling in neuroscience. Curr. Opin. Neurobiol. 31, 51–61 (2015).

PubMed

Google Scholar
Yeh, C.-H. et al. Cross-frequency coupling and intelligent neuromodulation. Cyborg. Bionic. Syst. 4, 0034 (2023).

PubMed
PubMed Central

Google Scholar
Tort, A. B. et al. Dynamic cross-frequency couplings of local field potential oscillations in rat striatum and hippocampus during performance of a T-maze task. Proc. Natl. Acad. Sci. USA 105, 20517–20522 (2008).

PubMed
PubMed Central

Google Scholar
Takeuchi, S. et al. Gamma oscillations and their cross-frequency coupling in the primate hippocampus during sleep. Sleep 38, 1085–1091 (2015).

PubMed
PubMed Central

Google Scholar
Muthuraman, M. et al. Cross-frequency coupling between gamma oscillations and deep brain stimulation frequency in Parkinson’s disease. Brain 143, 3393–3407 (2020).

PubMed
PubMed Central

Google Scholar
Motoi, H. et al. Phase-amplitude coupling between interictal high-frequency activity and slow waves in epilepsy surgery. Epilepsia 59, 1954–1965 (2018).

PubMed
PubMed Central

Google Scholar
Hyafil, A., Giraud, A.-L., Fontolan, L. & Gutkin, B. Neural cross-frequency coupling: connecting architectures, mechanisms, and functions. Trends Neurosci. 38, 725–740 (2015).

PubMed

Google Scholar
Fabus, M. S., Sleigh, J. W. & Warnaby, C. E. Effect of propofol on heart rate and its coupling to cortical slow waves in humans. Anesthesiology 140, 62 (2024).

PubMed

Google Scholar
Lechinger, J., Heib, D. P. J., Gruber, W., Schabus, M. & Klimesch, W. Heartbeat-related EEG amplitude and phase modulations from wakefulness to deep sleep: interactions with sleep spindles and slow oscillations. Psychophysiology 52, 1441–1450 (2015).

PubMed

Google Scholar
Faes, L., Marinazzo, D., Jurysta, F. & Nollo, G. Linear and non-linear brain–heart and brain–brain interactions during sleep. Physiol. Meas. 36, 683 (2015).

PubMed

Google Scholar
Dumont, M. et al. Interdependency between heart rate variability and sleep EEG: linear/non-linear?. Clin. Neurophysiol. 115, 2031–2040 (2004).

PubMed

Google Scholar
Candia-Rivera, D., Faes, L., de Vico Fallani, F. & Chavez, M. Measures and Models of Brain-Heart Interactions. IEEE Rev. Biomed. Eng. 14, 1–17 (2025).
Zhang, C., Yeh, C.-H. & Shi, W. Variational phase-amplitude coupling characterizes signatures of anterior cortex under emotional processing. IEEE J. Biomed. Health Inform. 27, 1935–1945 (2023).

PubMed

Google Scholar
Dang-Vu, T. T. et al. Spontaneous neural activity during human slow wave sleep. Proc. Natl. Acad. Sci. USA 105, 15160–15165 (2008).

PubMed
PubMed Central

Google Scholar
Svanborg, E. & Guilleminault, C. EEG frequency changes during sleep apneas. Sleep 19, 248–254 (1996).

PubMed

Google Scholar
Simon, C. W. & Emmons, W. H. EEG, consciousness, and sleep. Science 124, 1066–1069 (1956).

PubMed

Google Scholar
Hermann, B. et al. Aberrant brain–heart coupling is associated with the severity of post cardiac arrest brain injury. Ann. Clin. Transal. Neurol. 11, 866–882 (2024).

Google Scholar
Stein, P. K. & Pu, Y. Heart rate variability, sleep and sleep disorders. Sleep. Med. Rev. 16, 47–66 (2012).

PubMed

Google Scholar
Lyu, J., Shi, W., Zhang, C. & Yeh, C.-H. A novel sleep staging method based on EEG and ECG multimodal features combination. IEEE Trans. Neural Syst. Rehabil. Eng. 31, 4073–4084 (2023).

PubMed

Google Scholar
Kong, S. D. et al. Heart rate variability during slow wave sleep is linked to functional connectivity in the central autonomic network. Brain Commun. 5, fcad129 (2023).

PubMed
PubMed Central

Google Scholar
Zoccoli, G. & Amici, R. Sleep and autonomic nervous system. Curr. Opin. Physiol. 15, 128–133 (2020).

Google Scholar
Reyes del Paso, G. A., Langewitz, W., Mulder, L. J., Van Roon, A. & Duschek, S. The utility of low frequency heart rate variability as an index of sympathetic cardiac tone: a review with emphasis on a reanalysis of previous studies. Psychophysiology 50, 477–487 (2013).

PubMed

Google Scholar
Lee, H., Park, J., Kim, H. & Lee, K.-J. New rule-based algorithm for real-time detecting sleep apnea and hypopnea events using a nasal pressure signal. J. Med. Syst. 40, 1–12 (2016).

Google Scholar
Rolón, R. E. et al. Automatic scoring of apnea and hypopnea events using blood oxygen saturation signals. Biomed. Signal Process. 62, 102062 (2020).

Google Scholar
Eckert, D. J. & Malhotra, A. Pathophysiology of adult obstructive sleep apnea. Proc. Am. Thorac. Soc. 5, 144–153 (2008).

PubMed
PubMed Central

Google Scholar
De Chazal, P., Heneghan, C. & McNicholas, W. T. Multimodal detection of sleep apnoea using electrocardiogram and oximetry signals. Philos. Trans. R. Soc. A 367, 369–389 (2009).

Google Scholar
Berry, R. B. et al. Rules for scoring respiratory events in sleep: update of the 2007 AASM manual for the scoring of sleep and associated events: deliberations of the sleep apnea definitions task force of the American Academy of Sleep Medicine. J. Clin. Sleep. Med. 8, 597–619 (2012).

PubMed
PubMed Central

Google Scholar
Dingli, K. et al. Spectral oscillations of RR intervals in sleep apnoea/hypopnoea syndrome patients. Eur. Resp. J. 22, 943–950 (2003).

Google Scholar
Dissanayake, H. U. et al. Association between autonomic function and obstructive sleep apnea: a systematic review. Sleep. Med. Rev. 57, 101470 (2021).

PubMed

Google Scholar
Li, X. et al. Interaction effect of obstructive sleep apnea and periodic limb movements during sleep on heart rate variability. J. Sleep. Res. 28, e12861 (2019).

PubMed

Google Scholar
Candia-Rivera, D. & Cnrs, M. C. In 13^th Conference of the European Study Group on Cardiovascular Oscillations https://hal.science/hal-04813786v1/file/2024205034.pdf (ESGCO, 2024).
Faes, L. et al. Predictability decomposition detects the impairment of brain–heart dynamical networks during sleep disorders and their recovery with treatment. Philos. Trans. R. Soc. A 374, 20150177 (2016).

Google Scholar
Schreiner, T., Petzka, M., Staudigl, T. & Staresina, B. P. Respiration modulates sleep oscillations and memory reactivation in humans. Nat. Commun. 14, 8351 (2023).

PubMed
PubMed Central

Google Scholar
Tort, A. B., Hammer, M., Zhang, J., Brankačk, J. & Draguhn, A. Temporal relations between cortical network oscillations and breathing frequency during REM sleep. J. Neurosci. 41, 5229–5242 (2021).

PubMed
PubMed Central

Google Scholar
Tang, X. & Dworkin, B. R. Baroreflexes of the rat. VI. Sleep and responses to aortic nerve stimulation in the dmNTS. Am. J. Physiol. Regul. Integr. Comp. Physiol. 298, R1428–R1434 (2010).

PubMed
PubMed Central

Google Scholar
Conway, J., Boon, N., Jones, J. V. & Sleight, P. Involvement of the baroreceptor reflexes in the changes in blood pressure with sleep and mental arousal. Hypertension 5, 746–748 (1983).

PubMed

Google Scholar
Batini, C., Moruzzi, G., Palestini, M., Rossi, G. F. & Zanchetti, A. Persistent patterns of wakefulness in the pretrigeminal midpontine preparation. Science 128, 30–32 (1958).

PubMed

Google Scholar
Laguzzi, R., Reis, D. J. & Talman, W. T. Modulation of cardiovascular and electrocortical activity through serotonergic mechanisms in the nucleus tractus solitarius of the rat. Brain Res. 304, 321–328 (1984).

PubMed

Google Scholar
Magnes, J., Moruzzi, G. & Pompeiano, O. Synchronization of the EEG produced by low-frequency electrical stimulation of the region of the solitary tract. Arch. Ital. Biol. 99, 33–67 (1961).

Google Scholar
Wang, Y., Shi, W. & Yeh, C.-H. A novel measure of cardiopulmonary coupling during sleep based on the synchrosqueezing transform algorithm. IEEE J. Biomed. Health Inform. 27, 1790–1800 (2023).

PubMed

Google Scholar
McAllen, R. & Spyer, K. The baroreceptor input to cardiac vagal motoneurones. J. Physiol. 282, 365 (1978).

PubMed
PubMed Central

Google Scholar
Gilbey, M., Jordan, D., Richter, D. & Spyer, K. Synaptic mechanisms involved in the inspiratory modulation of vagal cardio-inhibitory neurones in the cat. J. Physiol. 356, 65–78 (1984).

PubMed
PubMed Central

Google Scholar
Moberly, A. H. et al. Olfactory inputs modulate respiration-related rhythmic activity in the prefrontal cortex and freezing behavior. Nat. Commun. 9, 1528 (2018).

PubMed
PubMed Central

Google Scholar
Bagur, S. et al. Breathing-driven prefrontal oscillations regulate maintenance of conditioned-fear evoked freezing independently of initiation. Nat. Commun. 12, 2605 (2021).

PubMed
PubMed Central

Google Scholar
Khalighi, S., Sousa, T., Santos, J. M. & Nunes, U. ISRUC-Sleep: a comprehensive public dataset for sleep researchers. Comput. Methods Prog. Biomed. 124, 180–192 (2016).

Google Scholar
O’reilly, C., Gosselin, N., Carrier, J. & Nielsen, T. Montreal Archive of Sleep Studies: an open-access resource for instrument benchmarking and exploratory research. J. Sleep. Res. 23, 628–635 (2014).

PubMed

Google Scholar
Berry, R. B. et al. AASM scoring manual updates for 2017 (version 2.4). J. Clin. Sleep. Med. 13, 665–666 (2017).

PubMed
PubMed Central

Google Scholar
Pan, J. & Tompkins, W. J. A real-time QRS detection algorithm. IEEE Trans. Biomed. Eng. 3, 230–236 (1985).
McSharry, P. E., Clifford, G. D., Tarassenko, L. & Smith, L. A. A dynamical model for generating synthetic electrocardiogram signals. IEEE Trans. Biomed. Eng. 50, 289–294 (2003).

PubMed

Google Scholar
Wu, Z. & Huang, N. E. Ensemble empirical mode decomposition: a noise-assisted data analysis method. Adv. Data Sci. Adapt. 1, 1–41 (2009).

Google Scholar
Huang, N. E. et al. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis. Proc. R. Soc. Lond. Ser. A Math. Phys. Eng. Sci. 454, 903–995 (1998).

Google Scholar
Wobbrock, J. O., Findlater, L., Gergle, D. & Higgins, J. J. In Proc. 11^th International Conference on Human Factors in Computing Systems. http://faculty.washington.edu/wobbrock/pubs/chi-11.06.pdf (SIGCHI, 2011).
Feys, J. Nonparametric tests for the interaction in two-way factorial designs using R. R. J. 8, 367–378 (2016).

Google Scholar

A novel holistic metric for sustainability assessment of photovoltaic/battery systems

In this section, the results of each sustainability metric are individually analysed to examine the dynamics of battery performance, PV power generation, and load behavior. Finally, the holistic metric is evaluated to provide a comprehensive overview of the cases studied and their overall resilience.

Battery sustainability index (BSI)

In this study, the BSI was designed to provide a computationally tractable representation of battery health by quantifying SOC excursions and cycling frequency via weightings supported by degradation trends in the literature^12,20. While this formulation captures the operational stress imposed on batteries, it does not include detailed electrochemical models such as Rainflow Counting or calendar aging estimation. For example, in Berlin, the BSI highlighted more frequent low-SOC events—reflecting deeper discharge cycles—which are known to accelerate degradation in both NMC and LFP batteries. These findings are consistent with trends reported in experimental investigations on hybrid energy storage, such as in²⁴, where lithium-ion battery degradation was significantly affected by real-world fluctuations in renewable energy output. The incorporation of such empirical modelling approaches and hybrid storage configurations will be a natural extension of this work, particularly to bridge the gap between operational patterns and electrochemical wear.

The SOC is maintained above the minimum threshold of 20% for most of the operational hours, as set by HOMER. However, the frequency and duration of excursions below the safe operating range vary significantly across locations (see Fig. 3). The SOCSI is calculated for all the study cases, as shown in Table 2.

For Cairo (a), the SOC profile indicates that the battery spends most of the year at or near full capacity (100% SOC). This suggests an oversized energy storage system, where PV generation consistently exceeds demand. The frequent operation at full SOC may indicate underutilization of the system’s capacity and potential for optimization to reduce oversizing.

For Berlin (b), the SOC profile shows more variation, with periods of SOC decreasing closer to the 20% threshold. However, the battery generally maintains the SOC within the safe range for a significant portion of the year. The balanced performance reflects a system that is reasonably well sized, although there may still be occasional mismatches between PV generation and demand.

For Riyadh (c), the SOC profile demonstrates the most consistent behavior, with the SOC remaining stable and within the safe range for most of the operational hours. The battery rarely exists at critical levels, which reflects effective PV generation and proper system sizing suited to Riyadh’s high solar irradiance.

The evaluation of the BSI and SOCSI over the three years (2017–2019) reveals significant differences in battery performance due to regional variations in solar energy availability, operational dynamics, and battery sizing, as shown in Table 3.

Table 3 Battery sustainability Indices.

Fig. 3

Comparison of state-of-charge (SOC) behavior across Cairo, Berlin, and Riyadh for 2017, illustrating the temporal dynamics and variability influenced by regional solar conditions and battery management.

Table 4 SOC quantitative indicators.

Although Fig. 3 provides insight into the SOC behavior over time, the SOC pattern is highly dynamic and influenced by several interdependent variables, including load demand fluctuations, solar radiation variability, and autonomy configuration. As such, determining the long-term sustainability or degradation risk of a battery on the basis of SOC plots or average values alone may lead to ambiguous conclusions. This complexity highlights the necessity of developing a comprehensive metric that encapsulates not only SOC trends but also their implications for battery longevity and system resilience.

Table 4 presents quantitative indicators of battery SOC behavior across three distinct locations (Cairo, Berlin, and Riyadh) over three years, including the average SOC, standard deviation, and cumulative time the battery operated outside the optimal SOC window. The results reveal significant spatial and temporal variations in battery operating patterns. For example, Cairo presented relatively high average SOC values (72.3–82.7%) with decreasing standard deviations over time (32.1–11.9%), reflecting tighter SOC control but frequent exposure to high SOC levels that risk accelerated degradation. Conversely, Berlin’s SOC remained generally lower (56.4–63.4%), yet with consistently high variability (approximately 34%), indicating less consistent SOC management. Riyadh displayed a stable average SOC (approximately 70%) and moderate standard deviation (≈ 20%), suggesting relatively balanced operation.

The SOCSI results reveal significant differences in battery performance across the studied locations, highlighting the importance of optimizing battery autonomy on the basis of regional solar conditions. In Cairo, an oversized PV system designed to provide one full day of autonomy consistently results in lower SOCSI values. This is because the system generates more power than the load demand does, causing the battery to remain near its maximum SOC for most of the year. While this setup reduces deep discharges, it limits the battery’s cycling within the optimal SOC range, resulting in less dynamic operation and higher degradation risks. Therefore, despite ensuring a reliable energy supply, Cairo’s battery system operates suboptimally in terms of long-term viability, with the BSI reflecting this imbalance between autonomy and SOC management.

However, these conventional statistical measures (mean and standard deviation) fail to capture critical aspects of battery sustainability. For example, both Berlin 2017 and Cairo 2017 show similar standard deviations (33.8% vs. 32.1%), yet the time spent outside the safe SOC range differs drastically (3461 h vs. 5408 h). This illustrates that traditional statistics provide limited insight into the actual stress imposed on a battery. The unsafe time metric highlights these hidden risks, emphasizing the need for a composite index such as the BSI, which integrates both the frequency and severity of SOC deviations alongside cycling behavior. The BSI thus offers a more meaningful and actionable indicator of battery health, supporting informed decisions about system operation and maintenance planning.

In contrast, Berlin experiences higher SOCSI values due to limited solar energy, which requires the battery to operate within a narrow SOC range. The battery spends much of the year at the lower SOC limit of 20%, especially in colder climates. However, the system’s ability to maintain this low SOC range leads to better resilience in terms of cycle life. Berlin’s higher BSI indicates a more favourable balance between cycling frequency and SOC management than does Cairo, where the oversized system limits cycling dynamics.

Riyadh benefits from consistently high solar availability, leading to more balanced and dynamic battery operation. The battery in Riyadh operates within the optimal SOC range more frequently, which promotes both longevity and reduced degradation. Riyadh’s high BSI values underscore its more sustainable battery performance, attributed to its optimal balance between SOC stability and efficient cycling.

The results emphasize the need for location-specific battery operation strategies. In Cairo, the oversized PV system should be optimized to avoid excessive SOC levels, as one day of autonomy is excessive for such sustainable high solar radiation. Conversely, in Berlin and Riyadh, one day of autonomy is more appropriate given the limited solar availability and consistent solar conditions, respectively. These findings highlight that autonomy should be carefully adjusted to local solar power performance, optimizing both battery performance and long-term viability across diverse environmental contexts.

To assess how different degradation mechanisms or battery chemistries influence the BSI, a sensitivity study was conducted where the weights were varied as follows:

Scenario A SOC-dominant scenario (ω₁ = 0.7, ω₂ = 0.3).
Scenario B Base case (ω₁ = 0.6, ω₂ = 0.4).
Scenario C Equal contribution (ω₁ = 0.5, ω₂ = 0.5).
Scenario D Cycle dominant (ω₁ = 0.4, ω₂ = 0.6).

Table 5 BSI values under each scenario.

As shown in Table 5, the analysis demonstrates that BSI is responsive to changes in degradation emphasis. For example, in Cairo in 2017, the BSI increased by more than 30% between the SOC-dominant (0.556) and cycle-dominant (0.730) scenarios. This variability highlights the importance of selecting appropriate weights to reflect the actual degradation mechanisms relevant to battery technology and the usage profile.

Lithium-ion chemistries such as NMC, which are widely used in advanced systems, are highly sensitive to SOC-related degradation, justifying higher SOCSI weightings (e.g., ω₁ = 0.6–0.7). In contrast, chemistries such as LFP or future solid-state batteries, which exhibit greater tolerance to high SOC, may benefit from recalibrated weights with greater emphasis on cycling. Similarly, systems with high-frequency shallow cycling may prioritize the cycle term to better capture wear patterns.

This flexibility enhances the applicability of BSI across diverse technologies and operating conditions. The sensitivity results also reinforce the need for careful, data-driven selection of weights rather than arbitrary choices, ensuring that the index meaningfully reflects the sustainability of the battery system.

PV power behavior

Figure 4 shows the daily ratio of served load to PV energy for three cities, i.e., Cairo, Berlin, and Riyadh, over 2017. Cairo consistently has a stable ratio below 1, indicating that PV energy production generally exceeds the load throughout the year. This pattern reflects the effective sizing of the PV system and favourable solar conditions, which minimize seasonal variability and ensure consistent energy availability. Berlin, in contrast, demonstrated significant fluctuations in the ratio, with multiple peaks exceeding 1. These spikes indicate periods where the served load surpasses PV energy production, particularly during winter months with lower solar irradiance. This variability highlights the challenges of relying solely on PV systems in regions with significant seasonal changes in solar availability, necessitating greater reliance on storage. Riyadh displays a pattern similar to that of Cairo, with ratios below 1 for most of the year, although it experiences slightly more variability. Occasional peaks suggest short-term mismatches between PV energy generation and load, but the overall stability reflects Riyadh’s high solar irradiance and well-sized PV system.

Table 6 summarizes the ERE values for the three locations over three years. The ERE evaluates the PV system’s reliability in delivering energy to meet load demands by calculating the average fraction of daily PV-generated energy utilized effectively. Cairo’s ERE values, ranging from 0.55 to 0.63, indicate a consistent and reliable energy supply where PV production typically exceeds load requirements. This reflects a surplus of PV energy, which is effectively utilized to serve the load, minimizing reliance on storage.

Although Berlin has the lowest solar irradiance among the three locations, it presented the highest ERE value in 2017 (0.878). This outcome is attributed not to higher energy availability but to better alignment between the PV system output and the local load profile, resulting in more efficient utilization of the available PV energy. The relatively moderate and consistent load demand in Berlin allows a larger portion of the PV-generated energy to be consumed directly, thereby increasing the ERE despite limited solar resources.

The fluctuation in Berlin’s ERE values across the years, particularly the noticeable drop in 2018 (0.732), can be attributed to seasonal variations in irradiance, which affect the degree of mismatch between generation and load. Unlike Cairo, where PV generation often exceeds load requirements year-round, Berlin’s performance is more sensitive to annual solar variability. These differences highlight the importance of system design and regional climate in determining the effective utilization of PV energy.

Riyadh has an ERE of approximately 0.74, indicating a well-balanced system where PV energy production and load demands align effectively. The consistency of these values highlights Riyadh’s stable solar resources and effective system sizing, which ensures high reliability in energy delivery across all three years. These observations emphasize the importance of tailoring PV systems and energy management strategies to specific regional conditions to optimize reliability and performance.

Table 6 Energy reliability efficiency.

These year-over-year trends underscore the importance of location-specific PV system design and energy management. While Cairo and Riyadh show consistent reliability, Berlin’s variability suggests the need for more robust storage or adaptive load management in regions with less predictable solar patterns.

Load behavior

The weighting factor was set to 3 during peak demand hours (08:00–16:00), where the load reached 200 kW, and 1 during other hours. This reflects periods of highest grid stress and system vulnerability, ensuring that the PDM metric prioritizes demand-matching performance during operationally critical times.

Cairo consistently has high weighted PDM values throughout the study period, with most values remaining above 0.6, as shown in Fig. 5-a. This indicates a well-balanced energy supply and demand, particularly during both peak and off-peak hours. The solar resource availability in Cairo, coupled with an effective EMS, supports stable performance.

Berlin shows a more volatile weighted PDM profile with frequent dips below 0.5 (see Fig. 5-b). This suggests challenges in meeting energy demand, particularly during peak load periods. The lower solar irradiance in Berlin likely impacts the PV performance, and the peak load weight (ω = 3) accentuates this mismatch during high-demand periods.

Riyadh’s weighted PDM values are relatively stable, remaining consistently above 0.5, with fewer extreme dips than those in Berlin. The region benefits from strong solar resources, enabling good performance, as shown in Fig. 5-c. However, occasional drops indicate periods of higher unmet loads, possibly during peak load periods or extreme weather conditions. Table 4 provides further insight into the long-term performance of the metric.

Cairo achieves the highest average PDM across all years. This reflects effective solar PV and EMS practices that ensure a balance between load demand and energy supply. Berlin consistently has the lowest PDM values, highlighting the impact of reduced solar resource availability. Seasonal fluctuations and the higher weight assigned to peak loads exacerbate the challenges in matching energy demand. Riyadh exhibits strong performance, with PDM values close to those of Cairo, as shown in Table 7. Its stable solar irradiance supports effective load balancing, although minor dips indicate areas for improvement in peak load management.

Holistic sustainability index

The radar chart, shown in Fig. 6, effectively highlights the comparative sustainability performance of Cairo, Berlin, and Riyadh across average values of three key metrics: BSI, ERE, and PDM. Riyadh has the strongest overall sustainability profile, leading to both BSI (0.81) and ERE (0.75), indicating robust battery longevity and efficient energy reliability (see Figure y). Cairo has a relatively high PDM score (0.78), reflecting excellent load matching performance, although its lower BSI (0.66) and ERE (0.60) suggest potential oversizing issues and opportunities for optimizing battery cycling and system efficiency. Berlin, while exhibiting the highest ERE (0.80), has the lowest PDM (0.53), indicating challenges in demand matching, likely due to variable solar availability and seasonal load fluctuations. The intermediate BSI value (0.72) for Berlin indicates moderate battery performance but suggests that system flexibility improvements, such as hybrid storage or renewable diversification, could further enhance sustainability. Overall, the radar chart underscores the distinct trade-offs in system design and operational efficiency among cities, emphasizing the need for tailored strategies to improve sustainability metrics on the basis of regional characteristics.

The final HM values for the three locations, i.e., Cairo, Berlin, and Riyadh, over three years are presented in Table 8.

Table 8 Holistic sustainability results.

The results of the HM analysis provide insights into the performance of the locations studied over three years. Riyadh achieves the highest HM values across all years, reflecting its optimal balance between battery performance, energy utilization, and load-matching efficiency. Its high solar availability ensures that the battery operates within the optimal SOC range more frequently, promoting longevity while maintaining efficient energy conversion and demand matching.

Berlin has moderate HM values because of its efficient cycling dynamics and relatively strong energy utilization ERE. However, its limited solar availability requires the battery to operate near the lower SOC range for extended periods, increasing stress and degradation risks. While Berlin’s energy management strategy effectively maximizes energy utilization under constrained conditions, improving battery stress management could enhance its overall performance.

Cairo, despite benefiting from an oversized PV system that ensures a reliable energy supply, records comparatively lower HM values. The oversized system results in prolonged periods of high SOC, reducing cycling dynamics and leading to inefficiencies in energy utilization and load matching. Adjusting the system design to better align with local demand patterns and solar availability could significantly improve the performance of Cairo’s HM.

The contrasting system behaviors observed in Cairo and Berlin highlight the need for region-specific optimization strategies. Cairo’s oversized PV system ensures high reliability but leads to suboptimal battery cycling, which lowers the overall HM by underutilizing the battery and potentially reducing its lifespan. Conversely, Berlin’s smaller PV capacity results in greater battery stress due to increased cycling but achieves a higher ERE because of more effective energy utilization. To improve system performance, it is recommended that Cairo’s battery capacity be optimized to increase battery cycling and extend battery longevity. For Berlin, integrating hybrid renewable energy systems, such as combining wind power with PV, could reduce battery stress while maintaining energy reliability. Future work will explore these hybrid configurations to further enhance system flexibility, resilience, and sustainability across diverse climatic regions.

The HM advances beyond traditional reliability-based indices such as the LMI by incorporating a multidimensional assessment of PV-battery system performance. Unlike the LMI, which primarily evaluates how well the load demand is met, the HM integrates the BSI, ERE, and PDM into a unified framework. This integration allows the HM to capture operational inefficiencies and long-term degradation risks that the LMI tends to overlook. A clear example is observed in the Cairo case, where the LMI indicated acceptable reliability performance due to high load coverage, yet the HM revealed a lower score of 0.66, indicating an oversizing issue and excessive battery cycling. This demonstrates HM’s ability to detect subtle but critical design flaws, offering a more insightful evaluation tool for system designers aiming to balance performance, longevity, and reliability.

These findings reinforce the importance of adopting a holistic approach to sustainability assessment, as the interaction between battery performance, energy utilization, and load matching varies significantly across locations. The proposed HM framework effectively captures these interactions, offering actionable insights for optimizing system design and operation to maximize resilience across diverse environmental and operational contexts.

Economic assessment of PV-battery systems

To enhance the practical relevance of the proposed HM, this section presents a basic economic analysis focused on the levelized cost of energy (LCOE) and battery replacement scheduling. The analysis links sustainability indicators, such as the BSI, to cost performance, highlighting the economic benefits of improved system design (Table 9).

Table 9 Economic assumptions and parameters.

Levelized cost of energy (LCOE)

The LCOE is calculated as:

$${text{LCOE }} = {text{ }}Sigma {text{ }}left( {{text{C}}_{{text{t}}} /left( {{text{1 }} + {text{ r}}} right)^{{text{t}}} } right)/Sigma {text{ }}left( {{text{E}}_{{text{t}}} /left( {{text{1 }} + {text{ r}}} right)^{{text{t}}} } right)$$

where C_t is the cost in year t (capital, replacement, O&M), E_t is the energy delivered to load in year, r is the discount rate (6%) and T is the project lifetime (25 years).

The battery replacement time

The battery replacement time in this study was estimated by linking the BSI results, which reflect the combined effects of SOC stability and cycle usage, to the expected consumption of the battery’s rated cycle life. To link the proposed BSI to battery economic life, we assume a linear relationship between the BSI and the replacement time. This is based on the premise that optimal SOC management and reduced cycling stress, reflected in higher BSI values, enable the battery to approach its maximum potential lifespan. For lithium-ion batteries, a typical maximum calendar life of 15 years is assumed under ideal operating conditions. Thus, the battery replacement time is estimated as:

$$T_{{replace}} = 15 times BSI$$

When this method is applied, the estimated battery replacement times for Cairo, Berlin, and Riyadh are approximately 9, 13, and 14 years, respectively. These estimates align with practical expectations for lithium-ion battery deployments in well-managed off-grid systems. This approach integrates the technical sustainability analysis with economic planning by translating the BSI into an expected replacement timeline. The economic analysis conservatively assumes a minimum of one replacement during a standard 25-year system lifetime to reflect calendar aging and real-world degradation mechanisms. This integration of sustainability results with replacement scheduling ensures that the proposed metric not only assesses technical health but also informs practical cost planning.

Table 10 Economic analysis results.

As shown in Table 10, this basic cost analysis reveals that systems with better battery sustainability (high BSI) not only perform better technically but also reduce long-term costs. In particular,

By linking HM to LCOE and the replacement frequency, this study provides a comprehensive view of both the technical and economic performance.

Continue Reading

August 16, 2025

‘A structural dependence on heavy industry’: can South Korea wean itself off fossil fuels? | Environment

South Korea’s vital statistics

On a cool early morning on South Korea’s east coast, Eunbin Kang pointed to a monument to a vanishing era.

The 2.1GW Samcheok Blue power plant which came online in South Korea in January looms out of the headlands above a beach made internationally famous by a K-pop album shoot. It is expected to emit 13m tonnes of CO₂ annually, while its lifespan could stretch beyond 2050, the year by which the country has pledged to reach carbon neutrality. The country was building coal-fired power plants, said Kang, an activist who heads the Youth Climate Emergency Action group and relocated to this city to oppose the facility, “even as the climate emergency demands an immediate halt to fossil fuel expansion”.

But Samcheok is not an outlier. It is a symbol of the stark climate contradiction at the heart of the world’s 12th largest economy, celebrated for its technological prowess in semiconductors and electric vehicle batteries, yet among the top ten worst global climate performers.

Despite South Korea’s impressive climate pledges to reach net zero by 2050 with a 40% reduction in emissions from 2018 levels by 2030, fossil fuels still dominate its energy mix: 60% of electricity comes from coal and gas, while renewables make up just 9%, a quarter of the OECD average of 34%.

The Samcheok Blue power plant, expected to emit 13m tonnes of CO2 annually. Photograph: Bloomberg/Getty Images

Monopoly strangling transition

At the heart of South Korea’s climate failure is an energy model based on a state monopoly and central planning. Korea Electric Power Corporation (Kepco), the state-owned energy company, controls transmission, distribution and retail, while its subsidiaries dominate generation, creating structural challenges for competitors. These include Korea South-East Power, Korea Western Power and four other generation subsidiaries that together operate the vast majority of the country’s coal, gas and nuclear power plants.

Meanwhile, renewable energy developers face an obstacle course of regulatory barriers. Until recently, windfarm developers had to obtain 28 different permits from multiple ministries in a bureaucratic maze which created years of delays and significantly increased project costs, making many otherwise viable developments financially unfeasible. Progress was made in early 2025 with the passage of a long-awaited bill aimed at streamlining approvals, although the law won’t take effect until 2026.

Grid connection remains another hurdle. While electricity demand has grown by 98% over the past two decades, the transmission network has expanded by just 26%, but attempts to expand the grid have led to bitter local conflicts.

In Miryang, South Gyeongsang province, the government tried to compel residents to sell up to clear space for transmission towers and people faced violent crackdowns during a six-year standoff. Currently, a dozen such projects are stalled in the country.

K-pop fans in Samcheok with banners calling for an end to the power plant owing to its negative impact on the environment. Photograph: Bloomberg/Getty Images

Protest banners against the Samcheok Blue were erected at Maengbang Beach, which residents fear will be ruined by the plant. Photograph: Bloomberg/Getty Images

In February 2025, the National Assembly passed a Power Grid Special Act aimed at expanding transmission. But civic groups warn the law reinforces the country’s decades-old top-down model of infrastructure development, removing what few safeguards remained around public consultation and environmental review.

“We fully acknowledge that renewable energy transition requires transmission lines,” says Kim Jeong-jin from Friends of the Earth in Dangjin, where one project faced more than 10 years of delays due to local opposition. “But the repeated conflicts arise because the electricity is not even for local use, yet it causes damage to our region without any regard for our voices.”

The country’s energy strategy is guided by the Basic Plan for Electricity Supply and Demand, a 15-year forecast revised every two years. But the framework, which dates back to the 1960s, still prioritises centralised, large-scale power generation – a model built for coal and nuclear, and fundamentally incompatible with today’s decentralised, flexible renewable technologies.

Graphic

Political volatility worsens the problem. Each five-year presidential term brings a policy reversal. For instance, in 2017, President Moon Jae-in announced a nuclear phase-out; his successor, the now disgraced ex-president Yoon Suk Yeol, reversed course five years later. This whiplash undermines any long-term planning for renewables – a problem faced by democracies around the world.

The consequences are stark. After Russia’s invasion of Ukraine sent fossil fuel prices soaring, Kepco incurred enormous losses. In 2022 alone, South Korea faced an extra 22tn won (£11.9 bn) in LNG power costs. Yet the government kept electricity prices artificially low, a political choice that pushed Kepco’s debt to a staggering 205tn won (£111bn) by 2024.

The former president Yoon Suk Yeol reversed the plan to phase out nuclear. Photograph: Anthony Wallace/AFP/Getty Images

Despite this crisis, meaningful reform remains elusive. This entrenched monopoly system has effectively blocked the clean energy transition, with independent renewable producers struggling to gain meaningful access to a market dominated by fossil fuel interests.

Carbon-intensive by design

More broadly, South Korea’s postwar rise relied on energy-intensive industries: steel, petrochemicals, shipbuilding and semiconductors.

“This structural dependency on heavy and chemical industries makes the energy transition extraordinarily difficult,” says Park Sangin, a professor of economics at Seoul National University. “These industries are deeply embedded in the country’s economic fabric and require vast amounts of stable, cheap electricity.”

Powerful chaebols, or family-controlled conglomerates like Posco, Samsung and Hyundai, exert outsized influence on national policy. Their operations are supported by an electricity market designed for industrial stability, not climate mitigation.

The Hyundai shipyard in Ulsan, South Korea. Conglomerates exert outsized influence on national policy. Photograph: Bloomberg/Getty Images

And the problem isn’t just domestic; South Korea also finances and provides the infrastructure for fossil fuels globally. South Korean shipbuilders dominate the global market for LNG carriers. Public financial institutions also bankroll overseas fossil fuel projects.

One that was recently approved, the Coral Norte gas project in Mozambique, is projected to emit 489m tonnes of CO₂ across its lifecycle. At the same time, South Korea has emerged as one of the world’s top importers of Russian fossil fuels, even as other nations cut ties.

“This financing directly contradicts [South] Korea’s climate targets and makes a mockery of the Paris Agreement,” says Dongjae Oh, the head of the gas team at Solutions for Our Climate (SFOC). “It exposes the country’s hypocrisy – adopting climate targets at home while funding climate destruction abroad.”

Even climate-friendly institutions continue backing fossil fuels. The National Pension Service (NPS), one of the world’s largest pension funds, remains a major investor in coal and gas projects, despite a 2021 “coal-free” declaration. Three and a half years after this announcement, NPS only finalised its coal divestment strategy in December 2024, with a timeline that will delay implementation for domestic assets until 2030.

Wolsong nuclear power plant in Gyeongju, South Korea. The country’s national energy plan still prioritises coal and nuclear power. Photograph: Bloomberg/Getty Images

Smoke rises from an industrial complex in Ulsan. South Korea’s largest polluters made over 475bn won from selling unused carbon credits. Photograph: Bloomberg/Getty Images

Meanwhile, South Korea’s market-based climate policies have failed to drive meaningful change. The emissions trading scheme (K-ETS) was supposed to put a price on carbon when it launched in 2015.

But the system, which hands out free allowances to the largest companies, has instead created perverse incentives, according to campaign group Plan 1.5. The group carried out an analysis and found that South Korea’s 10 largest polluters have made over 475bn won (£258bn) from selling unused carbon credits between 2015 and 2022. The system that was meant to make polluters pay has instead rewarded them.

Next generation fights back

There is growing awareness of a climate crisis as the country begins to experience increasingly severe weather. In 2023 46 people died in floods that displaced thousands. More recently, torrential rains have again caused at least 26 deaths, followed by a record-breaking heatwave.

In March this year devastating wildfires swept across more than 48,000 hectares (118,610 acres) – roughly 80% of the area of Seoul – killing 31 people and destroying thousands of homes. The country’s disaster chief described the situation as “a climate crisis unlike anything we’ve experienced before”.

The prime minister, Kim Min-seok, has described the climate crisis as “the new normal”.

An excavator on a barge near the site of the port under construction for Samcheok Blue. The country has described the climate crisis as like ‘nothing we’ve experienced before’. Photograph: Bloomberg/Getty Images

Now a new generation of South Koreans is challenging the status quo through legal action. In February, a group of children gathered outside Posco’s office in Seoul. Among them was 11-year-old Yoohyun Kim, the youngest plaintiff in a groundbreaking lawsuit against Posco.

The case aims to block the company’s plan to reline an old coal-fired blast furnace, a move that would extend its life by 15 years and emit an estimated 137m tonnes of CO₂.

“I came here during my precious winter break, my last as an elementary school student, because I want to protect all four seasons,” Yoohyun told supporters. “Spring and autumn are disappearing with climate change – and with them, the chance for children like me to play freely outside.”

The lawsuit is the first of its kind globally to target traditional blast furnace production. It follows a crucial ruling by South Korea’s constitutional court last August which found that the government’s climate policies violated the rights of future generations by failing to set legally binding targets for 2031-50.

In March, residents and activists filed another suit over the government’s approval of the world’s largest semiconductor cluster in Yongin, backed by a 360tn won (£195bn) Samsung investment. The suit argues that the project’s 10GW electricity demand and new LNG plants contradict climate regulations and corporate sustainability commitments.

A Kepco employee at work. The company is state-owned and has created structural problems for competitors. Photograph: Bloomberg/Getty Images

Kim Jeongduk, an activist from Political Mamas who participated in protests against the Samcheok Blue plant with her child, sees this as a generational struggle.

“Growing up in Pohang, I saw smokestacks fill the sky on my way to school every day. My throat would hurt from fine dust, and iron particles would collect on our windowsills,” she recalls.

“Adults always said: ‘Thanks to Posco, our region survives.’ I don’t want my child to grow up with that same false choice between a healthy environment and economic survival.”

The international data shows that South Korea’s emissions peaked in 2018, and have been falling, with a brief jump after Covid, ever since. The government maintains that it is making progress on its climate goals, although critics argue that it is relying on some wonky calculations around its 2030 emission reduction target, confusing net with gross emissions.

“South Korea is actively pursuing bold reduction of coal power generation through prohibiting new permits for coal power plants and phasing out ageing facilities,” the ministry said in a statement, arguing that any remaining coal plants operating beyond 2050, such as those approved before the 2021 ban, would be addressed through “carbon capture and storage technology and clean fuel conversion” in a way “not inconsistent with our carbon neutrality commitment”.

But independent analysis suggests these measures fall well short. “The Basic Plan has no specific plan for how to expand renewable energy,” says Prof Park. “There are vague targets, but no timeline, no locations. In stark contrast, the nuclear roadmap is extremely detailed and specific.”

His recent research using the Global Change Assessment Model shows the current plan would fall short of meeting South Korea’s 2030 emissions targets by approximately 6-7%.

A more ambitious policy focused on offshore wind expansion and a complete phase-out of coal by 2035 could not only meet climate goals but reduce power sector emissions by 82% by 2035.

Operations – ready-mixed concrete towers – at Ulsan port. Experts say there are no plans for the country to develop renewable. Photograph: Bloomberg/Getty Images

When confronted with criticisms of its emissions accounting, South Korea’s environment ministry defended its approach: “Our emissions reduction target calculation method considers international regulations and major country cases. Countries like Japan and Canada use similar calculation methods for their 2030 NDCs,” a spokesperson said.

The ministry added that although previous targets used the older 1996 IPCC guidelines, from 2024 they have begun using the updated 2006 standards for national greenhouse gas statistics.

Back in Samcheok, Eunbin Kang looks out at the coal plant that now dominates the coastal landscape.

“I dream of a society where exploitation and plunder are replaced by decentralisation and autonomy,” she says. “I want to contribute to spreading lifestyles and policies that allow everyone to lead a good life without requiring a lot of electricity or money.”

Continue Reading

August 16, 2025

Synergistic effects of commensals and phage predation in suppressing colonization by pathogenic Vibrio parahaemolyticus

Commensal intestinal bacteria can protect shrimp from the pathogenic Vibrio infection

To investigate the role of commensal bacteria in protecting shrimp from pathogenic Vibrio infections, we reconstructed a simplified, yet ecologically representative, version of the shrimp gut microbiota in vitro. Based on V3-V4 of 16S rRNA gene amplicon sequencing from healthy shrimp in our previous work¹⁹, we identified four dominant bacterial phyla including Proteobacteria, Bacteroidetes, Firmicutes, and Actinobacteria. From these groups, we curated a panel of twelve cultivable strains, each isolated from the healthy individual gut and taxonomically classified via whole-genome sequencing (see Methods) (Supplementary Table 1). These isolates formed a synthetic consortium, which we termed “Com12”, designed to capture key ecological and functional diversity of the native microbiota.

To examine interactions between commensal and pathogenic bacteria, we conducted in vitro co-culture experiments with Com12 consortium and two Vibrio strains: one pathogenic (Vibrio parahaemolyticus strain VP6) and one putatively beneficial (Vibrio spp. VA3). Initially, taxonomic characterization of these two Vibrio strains was performed via whole-genome sequencing, and phenotypic classification was based on the presence of virulence-associated genes and mortality assays. Genome annotation revealed that VP6 carries the pirAB toxin gene, whereas VA3 lacks this virulence determinant, a finding further confirmed by genetic analysis (Supplementary Fig. 1). Infection experiments demonstrated that VP6 induced significant mortality and vibriosis-specific symptoms, yet shrimp exposed to VA3 exhibited no detectable difference in survival relative to the control group (Fig. 1a, Supplementary Fig. 2). Co-culture experiments further revealed that both Vibrio strains significantly altered the structure of the Com12 consortium, with each strain becoming dominant within the microbial community (Supplementary Fig. 3). Notably, VA3 exhibited a distinct growth advantage over VP6 in Com12, suggesting it may competitively inhibit VP6 colonization.

Fig. 1: Characterization of commensal and pathogenic Vibrio strain in shrimp.

a Experimental design schematic. In the invasion assay (left), shrimp maintained under standard aquaculture conditions were exposed to pathogen V. parahaemolyticus VP6 or commensal Vibrio VA3. In the alternative invasion assay (right), shrimp pretreated with antibiotic cocktail (see “Methods“) and were randomized into groups with different bacterial strains in seawater containing (i) Com12 strains, (ii) individual strains, or (iii) combinatorial treatments with the pathogen (VP6). b Survival rate curves of shrimp exposed to Vibrio strains (VP6 and VA3). Fresh bacterial cultures (5 × 10^6 CFU/mL) were added to the shrimp (n = 20 per group) housing water. Untreated shrimp served as control. Black dashed lines indicate median survival. Statistics: log-rank test (P < 0.0001). c Survival rate curves of antibiotic-treated shrimp (n = 20 per group). Groups: Control (antibiotics only), VP6 (antibiotic-treated shrimp exposed to VP6), VP6+Com12 group (antibiotic-treated shrimp exposed to VP6 and the Com12 consortium), and VP6+Com12 + VA3 group (antibiotic-treated shrimp exposed to VP6, the Com12 consortium and VA3). Black dashed lines indicate median survival. Statistics: log-rank test (P values shown or P < 0.0001).

To further assess the protective role of VA3, we pretreated shrimp with antibiotics to minimize the native microbiota and subsequently recolonized some individuals with the Com12 consortium, −+ VA3, at a concentration of 5 × 10^6 CFU/mL, a dosage informed by previous studies^20,21. All shrimp were then challenged with VP6 (5 × 10^6 CFU/mL). By day 5, the VP6-only group exhibited 100% mortality (Fig. 1b). In contrast, both the Com12 + VA3 and Com12-only groups showed increased survival, with the former achieving a significantly higher survival rate (69%) compared to the Com12-only group (49%, P < 0.05; Fig. 1b). These results highlight the role of VA3 in enhancing the Com12-based resistance to VP6 infection. However, further investigations are needed to determine whether VA3 alone is sufficient to confer protection in the absence of Com12 − an open question that warrants deeper exploration in future studies.

In addition, to explore the potential of phage therapy in augmenting microbiota-mediated colonization resistance, we isolated an obligate lytic myovirus, VP6phageC, using VP6 as the host (Supplementary Fig. 4a, b). The host range of VP6phageC was assessed against each monospecies within the Com12 consortium, including VA3, using phage infection assays²². As expected, VP6phageC failed to form plaques on any Com12 members and VA3 (Supplementary Table 2), confirming its strict specificity for VP6. This specificity makes VP6phageC as a promising candidate for investigating whether a combined approach—leveraging phage predation alongside microbiota modulation, can further enhance shrimp defenses against VP6 colonization.

Commensal bacteria and phage can supress Vibrio pathogen growth

To investigate the interplay between commensal bacteria and phage in resisting Vibrio pathogen invasion, we employed the Com12 consortium as a model system to simulate pathogen invasion in vitro and evaluate the combined effects of commensal-derived colonization resistance and phage predation. These experiments were performed on a 96-well microplate platform, using time-resolved measurements of mono-species to track the dynamic changes over 48-h period (Fig. 2a, see “Methods”).

**Fig. 2: Temporal dynamics of the synthetic Com12 consortium composition under different treatments.**

Commensal-mediated resistance was evaluated by co-culturing Com12 and Com12 + VA3 with VP6 and monitoring strain abundance over time. VP6 rapidly dominated Com12, reaching 70% relative abundance within 6 h before stabilizing (Fig. 2b). However, the presence of VA3 significantly restricted VP6 expansion, limiting its abundance to 15%. Phage addition further suppressed VP6 to 5%, and the combination of VA3 and phage nearly eliminated VP6, reducing its abundance to less than 1% (Fig. 2c).

Further analysis of the co-culture dynamics revealed significant shifts in both bacterial abundance and overall community diversity. The introduction of VA3, phage, or their combination led to a marked reduced in VP6 abundance, accompanied by increased abundance of other strains within the Com12 consortium (Fig. 2d). To explore how VP6 suppression relates to overall community structure and community diversity, we performed linear regression analyses between the relative abundance of VP6 and community diversity, measured using the Shannon index. Importantly, the diversity metric included all community members, including VA3 and VP6, to reflect the total ecological outcome under each treatment condition. In the VA3-only treatment, the correlation between VP6 abundance and diversity was weak and statistically non-significant (R-square = 0.03, F-statistics = 0.51, P = 0.486). In contrast, phage treatment showed a positive correlation with diversity (R-square = 0.60, F-statistics = 24.30, P < 0.001). When VA3 and phage were combined, while the correlation between VP6 abundance and diversity was reduced, the positive relationship between diversity and pathogen suppression was maintained (R-square = 0.68, F-statistics = 33.30, P < 0.001) (Fig. 2d).

These results demonstrate that while Com12 alone impose a threshold on VP6 colonization, VA3 and phage act as potent inhibitors, with their combined application synergistically enhance colonization resistance. Importantly, phage contributed to increase microbial diversity, whereas VA3 appears to influence pathogen abundance without directly altering diversity, suggesting complementary mechanisms in pathogen exclusion. Together, these results underscore the potential of integrating commensals and phages as strategy to fortify maintaining microbiome stability and prevent pathogen invasion.

Timing of commensal and phage administration is critial for effective colonization resistance

In our study of the Com12 consortium, we identified a priority effect that influenced pathogen invasion, particularly when VP6 was introduced at different stages of Com12 growth. To evaluate how phage treatment could restore the consortium’s resistance following VP6 invasion, we reconducted co-culture experiments where Com12 was exposed to VP6, and VP6phageC (MOI = 1) was introduced at various time points (see “Methods”). Our findings indicated that the timing of phage introduction significantly affected its efficacy (Fig. 3a). More specifically, when VP6 was co-cultured with Com12 for 6 h or more before the addition of phage, the suppressive effect of the phage was notably diminished, with VP6 relative abundance surged to 70% of the community, suggesting that phage-mediated suppression was less effective after this time window (Fig. 3b). A comparison of absolute VP6 concentrations using copy numbers (unless otherwise specified) in the consortium further corroborated this observation, showing a marginal but not statistically significant reduction in VP6 when phage was added after 6 h (P > 0.05) (Fig. 3c). These results suggested that the efficacy of phage treatment is compromised once VP6 has had a chance to establish itself within the consortium for an extended period.

**Fig. 3: Timing-dependent efficacy of combinatorial interventions against pathogen VP6 invasion.**

Next, we explored the potential for combining certain commensal species with phage-specific predation to further bolster colonization resistance. Specifically, we assessed the effects of introducing commensal VA3, alongside the lytic phage VP6phageC. In these experiments, VA3 was introduced into Com12 consortium following VP6 invasion, with concurrent VP6phageC treatment. Monitoring of the consortium dynamics revealed significant inhibition of VP6 by this combined treatment. While VA3 alone significantly reduced the relative abundance of VP6 to 15% and its biomass to 5×10^7 CFU/mL within 48 h (Fig. 2c), the combined treatment of VA3 and VP6phageC further amplified this inhibitory effect. In this dual intervention, VP6 proliferation was almost entirely eradicated, with its relative abundance dropping below 1% and biomass reduced to less than 1 × 10^6 CFU/mL (Fig. 3d, e). Even when phage was introduced after a 6-h delay, both VP6 relative abundance and biomass remained significantly lower compared to treatments using either VA3 or phage alone (P < 0.001) (Fig. 3d, e). These results underscore that although phage-mediated pathogen suppression is highly time-dependent, its effectiveness can be synergistically enhanced when combined with commensal bacteria such as VA3, which together provide robust colonization resistance and protect the microbiota from pathogen invasion.

To further investigate whether the Com12 consortium itself possesses intrinsic, self-regulated resistance to pathogen invasion, we conducted an additional experiment by introducing VP6 at various stages of Com12 growth. Relative abundance analyses revealed a notable difference in outcomes depending on the timing of VP6 introduction. When VP6 was co-cultured with Com12 from the start (0 h), it quickly dominated the consortium as reflected in its high relative abundance and biomass (Fig. 2b, upper right). In contrast, when VP6 was introduced after 2 h of Com12 growth, its proliferation was irreversibly inhibited, with VP6’s absolute concentration falling to less than 1 × 10^5 CFU/mL, and its relative abundance dropped below 1% (Fig. 3f, g).

Overall, these findings underscore a crucial, timing-dependent trait of colonization resistance within the consortium, suggesting that early establishment of the commensal consortium provided a robust barrier against pathogen invasion, emphasizing the importance of microbial community maturation. Conversely, when the pathogen was allowed to establish dominance before the consortium had fully matured, the protective capacity was significantly compromised.

Commensal bacteria supress pathogen by nutrient competition and prophage induction

To explore the potential mechanisms underlying the observed colonization resistance in above co-culturing experiments (Figs. 2, 3), we investigated pairwise interactions among members of the consortium Com12, including V. parahaemolyticus (VP6) and Vibrio spp.(VA3), using a conditional coculturing approach²³. In this experiment, each species was grown in cell-free spent media collected from other species, supplemented with 60% full-nutrient Marine Broth (2216MB). This experimental design eliminated direct cell-cell contact as a potential mechanism for colonization resistance, allowing us to focus solely on metabolic-mediated interactions.

Analysis of the growth rate and maximum biomass (OD600) of each species grown in spent media from other community members versus self-derived media revealed a negative correlation, although the absolute correlation index was less than 1 (Estimate = -0.51, R-square = 0.12, F-statistic = 13.13, P = 2.57e–06) (Fig. 4a, left; Supplementary Fig. 5). This suggests that most interspecies interaction were inhibitory, albeit the extent of inhibition primarily affected total biomass rather than growth rate. When examining the effects of other species in Com12 on VP6, we observed a similar inhibitory trend, but with a stronger negative correlation (Estimate = –1.31, R-square = 0.54, F-statistic = 23.60, P = 0.004) (Fig. 4a, right). Ranking the effects of others on VP6 showed that the inhibitory effects were largely mediated by VA3, particularly when considering growth rate independently of biomass (Fig. 4a, right).

**Fig. 4: Characterization of the interactions between commensals, VP6 and VA3.**

To further explore the interactions underlying these observations, we employed genome-scale metabolic modeling to assess the functional similarity between VP6 and each member of Com12, including VA3, based on protein composition overlap. Specifically, we quantified the proportion of protein families carried by VP6 that were also shared in each commensal (see “Methods”). Our results highlight that VA3, Rueg, and Tena as key contributors to protein-family overlap with VP6, suggesting their potential role in shaping VP6’s growth dynamics by nutrient competition (Fig. 4b). Notably, VA3, belonging to the same bacterial genus as VP6, exhibited the highest degree of overlap, reinforcing its potential for strong competitive interactions.

To experimentally validate the pairwise interaction between VA3 and VP6, we conducted direct growth competition assays, in which a 1:1 mixture of VA3 and VP6 cells was co-cultured in nutrient-rich medium (2216MB) for 48 h, alongside monoculture controls. VA3 displayed robust growth, achieving cell densities comparable to its monoculture controls (P = 0.52) (Fig. 4c). In contrast, VP6 growth was severely impaired in the presence of VA3, with its cell density significantly reduced by 2- to 8-fold (P = 0.004) during the 48-h competition period (Fig. 4c). Furthermore, VA3 consistently outcompeted VP6 in both total biomass (0.97 vs 0.75) and growth rate (0.13 vs 0.10, by hour) (Fig. 4c, Supplementary Fig. 6). Together, these results align with the suppression observed in Com12 upon the introduction of VA3 (Figs. 2, 3), reinforcing the notion that VA3 inhibits VP6 proliferation. The observed suppression appears to be primarily due to nutrient competition, particularly the overlap in nutrient utilization between the two strains.

A recent study revealed that prophages in Vibrio strains are inducible and play critical roles in strain competition within marine environments²⁴. Inspired by this, we identified two intact prophages in the genome of Vibrio VP6. Under nutrient-limited conditions, one prophage was highly induced, as evidenced by increased read depth in its corresponding genomic region (Fig. 4d). Transmission electron microscopy (TEM) of the filtered supernatant confirmed the presence of filamentous phage particles characteristic of Inoviruses, measuring approximately 1,800–2,000 nm in length and 5 nm in width (Fig. 4d, upper panel). The genome of this phage, termed Vpp2, closely resembled that of filamentous phages based on its size (10,298 bp) and gene annotation (Fig. 4d, lower panel).

To investigate the role of prophage Vpp2 in microbiota interaction, we assessed its induction in conditioned media from 13 donor strains, using media from VP6 as a control. Vpp2 production increased in all conditioned media except that from Deme (Fig. 4e, f). When nutrient-deficient SM buffer was used, Vpp2 production also increased, indicating that nutrient limitation is a key trigger for its activation (Fig. 4e). Co-culturing VP6 with VA3 or Com12 separately revealed continuous induction of Vpp2 over 48 h with VA3, whereas the induction was less pronounced with Com12 (Supplementary Fig. 7). Together, these findings indicate that both Com12 and commensal VA3 promote prophage induction in VP6, with VA3 exhibiting the most robust effect. This induction likely involves nutrient competition, which activates a stress response in VP6, suggesting that prophage activation may be part of a broader ecological strategy that influences the growth dynamics of VP6.

Synergistic interaction of commensal microbes and phage confers colonization resistance against pathogenic Vibrio in shrimp

The dynamics of the Com12 consortium, −+VA3, revealed that individual strains contribute variably to colonization resistance against pathogen invasion, with some strains playing more pivotal roles. To evaluate the protective capacity of each commensal, we further assessed shrimp survival following exposure to pathogen VP6 (Fig. 4g). Shrimp were pretreated with antibiotics as before (Fig. 2g) to minimize the influence of indigenous bacteria and then immersed in cultures of each strain (5×10^6 CFU/mL) or combinations, including Com12 and VA3, prior to VP6 exposure. With this assay system, we could rank the strains based on their abilities to protect shrimp from VP6 infection. Shrimp survival rates were significantly higher when pretreated with VA3 ( ~ 69.0%), Psyc ( ~ 44.0%), Rueg ( ~ 43.0%), or Halo ( ~ 38.0%) compared to the positive control group exposed only to VP6 ( ~ 23.0%). Other strains showed insufficient or adverse effects on survival.

To evaluate whether the subset consortium comprising VA3, Psyc, Rueg, and Halo (Com4) could protect shrimp from VP6 infection and whether this protective effect could be enhanced by phage addition within the complex intestinal microbiome, we let shrimp be colonized with Com4 (5×10^6 CFU/mL per strain) before exposing to VP6 (5×10^6 CFU/mL) (Fig. 5a). Then, the shrimp were maintained under standard aquaculture conditions and fed daily with phage ( ~ 10^9 PFU/g) throughout the experiment. Successful colonization by VP6 causes an acute infection over a five-day period with white hepatopancreas and empty digestive tracts (Supplementary Fig. 8), which is a typical symptom of vibriosis^25,26, whereas shrimp with Com4 could rapidly succumb to the infection.

**Fig. 5: Synergistic effects of phage and commensal strains on intestinal microbiota and VP6 in shrimp.**

While mortality occurred across all groups following VP6 exposure, the cumulative survival rate of shrimp in the Com4-phage treatment group significantly increased to 58% (P < 0.001) compared to >20% in the VP6-only challenge group (Positive control) (Fig. 5b). The survival rate in the Com4-phage group was also notably higher than in the phage-only treatment group, confirming in vitro findings (Fig. 3) that the combination of commensal bacteria and phage more effectively inhibits VP6 invasion.

In addition, both the phage and the Com4-phage treatments effectively suppressed VP6 colonization in the aquatic environment. Plate counting assays revealed that VP6 became nearly under detectable in water surrounding the shrimp after three days in the Com4-phage treatment group, whereas similar suppression was observed only on the fifth day in the phage-only group (Fig. 5c). Quantitative analysis of VP6 in shrimp intestinal samples collected on day 3 and 5 showed a similar trend: VP6 abundance significantly decreased by over 90% in both treatment groups compared to the positive group (Fig. 5d).

Interestingly, the phage (VP6phageC) and its susceptible Vibrio target (VP6) coexisted in the shrimp intestine of the phage-only treatment group. While most Vibrio strains isolated from gut samples remained susceptible to the wild type phage (Supplementary Fig. 9), the observed increased in phage particles alongside a decrease in Vibrio load suggests that phage-mediated suppression of VP6 was effective but limited within the intestinal environment.

To investigate the effects of the treatment on the shrimp gut microbiome, samples were collected at three time points for further analysis. Alpha diversity, as measured by the Shannon index, decreased in the positive control group but increased in the phage-only and Com4-phage treatment groups (Fig. 5f). Characterization of the microbiome revealed that both Com4 strains and VP6 successfully colonized the shrimp gut (Fig. 5g). Importantly, the relative abundance of VP6 was significantly lower in the Com4-phage treatment group compared to the positive control group and phage-only groups, demonstrating superior pathogen resistance and microbiome recovery in the Com4-phage treatment group.

A ternary plot of bacterial OTUs in shrimp gut samples demonstrated that the four commensal strains, in combination with phage predation, effectively suppressed VP6 colonization (Supplementary Fig. 10). These findings suggest that Com4 strains provide substantial resistance to pathogen colonization in the shrimp gut, complementing the inhibitory effects of phage. Co-occurrence network analysis of the gut microbiomes in the Com4-phage group revealed positive interactions between VP6 and other Vibrio species, including VA3 (Supplementary Fig. 11) This suggests that VP6, VA3, and indigenous Vibrio spp. occupy similar ecological niches within the shrimpgut, potentially contributing to complex competitive dynamics.

Together, these results highlight the synergistic effects of commensal bacteria and phage in enhancing colonization resistance against VP6. The combination of Com4 strains and phage not only improved shrimp survival rates but also restored microbiome diversity and reduced VP6 colonization more effectively than phage treatment alone. This underscores the potential of leveraging commensal-phage synergies to protect aquaculture species from pathogenic infections.

Continue Reading

August 16, 2025

Anthropic’s Claude AI Ends Harmful Chats Automatically

In a landmark move toward ethical AI development, Anthropic has enhanced its Claude Opus 4 and Opus 4.1 models with the ability to autonomously end harmful or unproductive conversations. This feature, part of the company’s model welfare initiative, represents a new frontier in self-regulating AI behavior.

AI Welfare: When the Model Walks Away

Anthropic’s research shows Claude AI can now recognize when a conversation repeatedly violates policy or includes toxic inputs. In such cases, the model disengages without human prompting, reducing risks of misalignment or fatigue and mirroring an emotional safeguard seen in humans facing abusive situations.

The company frames “model welfare” as a growing field, ensuring that AI systems have internal guidelines to handle stress or misuse, rather than relying solely on external filtering systems.

A Measured Advance in AI Safety

This functionality is carefully constrained. It only activates during a rare subset of disruptive interactions, such as persistent extreme profanity or ethical contradiction in user prompts. The goal is not to disrupt normal usage but to proactively shield the model from potentially damaging scenarios.

Critics Raise Important Questions

While praised for prioritizing safety, this innovation has sparked debate. Critics warn that if the model ends conversations too readily, it could limit legitimate dialogue or introduce unfair bias. Others point to deeper concerns: might an AI with this power develop expectations or “internal goals” of its own?

The Bigger Picture in AI Regulation

Anthropic’s development aligns with broader trends in AI ethics. The company also pioneered “preventative steering,” a safety training method injecting “undesirable trait vectors” like toxicity during fine-tuning to boost resilience in models. This and Claude’s new self-ending feature work together to promote robust and responsible AI behavior.

Continue Reading

August 16, 2025

You Can Drive Home a Peugeot 2008 in 18 Easy Payments

Lucky Motor Corporation has opened bookings for the Peugeot 2008, a locally assembled compact SUV now available in two variants, Active and Allure, with flexible installment options.

Features and Equipment

Both variants are powered by a 1.2-liter turbocharged PureTech engine producing 131 hp and 230 Nm of torque, paired with a 6-speed automatic transmission. The Active trim includes essential features such as LED daytime running lights, rear parking sensors, cruise control, and a 7-inch touchscreen infotainment system supporting Apple CarPlay and Android Auto.

The Allure variant offers additional features, including a panoramic sunroof, six airbags, automatic climate control, blind-spot monitoring, lane-keeping assist, and a 10-inch digital instrument cluster. Both models come equipped with Peugeot’s signature 3D i-Cockpit design.

The company has introduced a limited-time installment plan with 50% down payment and monthly installments over 18 months. This offer is available through select dealerships, including Peugeot Metropolis in Islamabad.

Pricing and Payment Schedule

Variant	Price (PKR)	50% Down Payment	18 Monthly Installments
Active	7,249,000	3,624,500	201,362
Allure	8,049,000	4,024,500	223,584

Terms and conditions apply. Prices may vary depending on the dealership and availability.

Availability

Bookings are now open nationwide. The installment plan is designed to make ownership more accessible amid rising vehicle costs. Deliveries will be managed based on stock availability at the time of booking.

Continue Reading

August 16, 2025

Engine Capital takes a stake in Avantor. Activist sees several ways to create value

Company: Avantor (AVTR)

Business: Avantor is a life science tools company and global provider of mission-critical products and services to the life sciences and advanced technology industries. The company’s segments include laboratory solutions and bioscience production. Within its segments, it sells materials and consumables, equipment and instrumentation and services and specialty procurement to customers in the biopharma and health care, education and government and advanced technologies and applied materials industries. Materials and consumables include ultra-high purity chemicals and reagents, lab products and supplies, highly specialized formulated silicone materials, customized excipients and others. Equipment and instrumentation include filtration systems, virus inactivation systems, incubators, analytical instruments and others. Services and specialty procurement include onsite lab and production, equipment, procurement and sourcing and biopharmaceutical material scale-up and development services.

Stock market value: $8.85 billion ($12.98 per share)

Activist: Engine Capital

Ownership: ~3%

Average Cost: n/a

Activist Commentary: Engine Capital is an experienced activist investor led by Managing Partner Arnaud Ajdler. He is a former partner and senior managing director at Crescendo Partners. Engine’s history is to send letters and/or nominate directors but settle rather quickly.

What’s happening

On Aug. 11, Engine sent a letter calling on Avantor’s board to focus on commercial and operational excellence, demonstrate organic growth, reduce costs, optimize the portfolio, refresh the board and use free cash flow to repurchase stock. Engine noted that the company can alternatively consider a sale.

Behind the scenes

Avantor is a market leading distributor of life science tools and products for the life sciences and advanced technology industries. The company is comprised of two segments: laboratory solutions (LSS) (67% of revenue) and bioscience production (BPS) (33% of revenue). LSS is one of the three top life sciences distributors in the world (Thermo Fisher and Merck KGaA being the other two).

BPS is a supplier of high-purity materials and is the leading supplier of medical-grade silicones. Despite being one of the few scaled global life science tool distribution platforms, the company has vastly underperformed. At its 2021 investor day, management projected earnings per share above $2 for 2025; and at its 2023 investor day, management targeted an EBITDA margin exceeding 20%. Now in 2025, these currently stand at 96 cents per share and 11.8%, respectively. Consequently, Avantor’s share price has declined 53.96%, 59.69%, and 43.41% over the past 1-, 3- and 5-year periods, as of Engine’s announcement Monday.

Engine believes that Avantor’s significant underperformance is a consequence of self-inflicted mistakes rooted in a flawed leadership team and framework. A complex matrix organizational structure and resultant lack of accountability have led to mass leadership turnover, including Avantor’s CEO, CFO and both segment leaders within the past three years, contributing to a dysfunctional decision-making process and inefficient employee structure.

The biggest casualty of this rocky management team is LSS, which has lost significant profitability and market share to its peers. Specifically, poor capital allocation decisions have destroyed significant value. In 2020 and 2021, Avantor spent a total of $3.8 billion to acquire Ritter, Masterflex and RIM Bio – companies that were notably purchased during the peak of the pandemic when life sciences businesses were trading at exceptionally high multiples. Applying Avantor’s next 12 months 10x multiple to the 28x average acquisition price implies over $2.4 billion in lost value on these acquisitions, contributing to the company’s high leverage.

On top of that, despite LSS’s ongoing underperformance and the need for strong leadership, from June 2024 to April 2025, LSS was left without a leader due to a non-compete lawsuit involving the hiring of its new segment leader, underscoring the operational dysfunction that has been taking place at the company.

But perhaps the nail in the coffin for this management team and board is that despite this cascading set of errors and the internal knowledge of these forecasted losses, they were still given a way out. In 2023, the company was approached by Ingersoll Rand to be acquired at an estimated $25-$28 per share, a 20%-35% premium of the share price at the time, yet the board inexplicably rebuffed this approach. Today, Avantor trades at just under $13 per share.

Enter Engine, who has announced an approximately 3% position in Avantor and is urging the board to focus the organization on commercial and operational excellence, demonstrate organic growth, reduce costs, optimize the portfolio, refresh the board and use free cash flow to repurchase its own stock.

Engine points out that Avantor’s reported $6.8 billion in revenue was stretched across 6 million stock keeping units, while Thermo’s peer segment achieves similar revenue with less than half the SKUs, indicating a large opportunity, specifically within LSS, to optimize the portfolio by concentrating purchases to improve inventory turns, rebates and margins.

Divesting non-core assets is another way to optimize the portfolio. For BPS, certain facilities operate in periods of extended downtime, limiting growth. For LSS, subscale facilities in smaller geographies may be more valuable to a competitor, and the same goes for some of the assets purchased under Avantor’s aforementioned acquisition spree.

On the cost discipline side, Avantor’s history of poor M&A and its low valuation should limit its accretive M&A opportunities, and while the company is on the path to reduce leverage below 3x, the market remains concerned that once this is achieved, they will simply resume this costly M&A strategy. Engine argues that free cash flow should instead be allocated evenly to share repurchases and debt reduction.

Additionally, executive compensation is also a concern. In 2024, despite organic revenue declining by 2% and a 7% share price decline, the board awarded CEO Michael Stubblefield 110% of his target annual bonus, underscoring the need to align these management incentives with shareholder value creation.

Engine believes that all of these changes would be best implemented with a comprehensive board refreshment. Adding directors with executive leadership, capital allocation, and distribution expertise to replace board members that have overseen years of value destruction, likely targeting chairman Jonathan Peacock specifically, should signal to the market the start of a new chapter. Engine believes that if these changes are properly implemented that Avantor shares would be worth between $22 and $26 per share by the end of 2027.

As a secondary option, Engine suggests that if a standalone path does not appear viable then the board should consider selling the entire company or splitting LSS and BPS into separate entities.

When Avantor acquired VWR, which is now the core of the LSS business, it was valued at about 12x EBITDA, or $6.5 billion, and BPS peers trade at a median of 17x EBITDA. Neither of these businesses’ valuations correspond to what Avantor trades at, roughly 8x EBITDA, and it’s possible that a strategic path could become the best way to unlock this value on a risk-adjusted basis. If this were to become the case, there is likely to be both private and strategic interest. New Mountain Capital previously owned Avantor prior to its IPO and still maintains an approximately 2% position. Strategics, like Ingersoll, would likely be interested as well, especially at a significant discount to what they once offered. Engine believes that Avantor could sell between $17 to $19 per share.

Overall, Engine makes not only a compelling case that major change is needed at Avantor, but also a clear multipath plan forward. While some of these changes are already underway: a new CEO is set to start next week and management announced a $400 million cost-cutting initiative, the sheer volume of change required here is unlikely to occur by Engine’s 2027 estimate.

Engine’s plan includes strengthening execution, instilling a culture of cost discipline, improving capital allocation, evaluating the company’s portfolio, aligning executive compensation to shareholder value creation and refreshing the board. Engine’s plan is the right one, but this is a company whose top line and operating margins have been in decline since 2022 and refreshing a board, instilling a new culture, reversing declining revenue and operating margins and evaluating and executing asset sales, many of which cannot be done simultaneously, is something that will likely take much longer than two years, particularly with the director nomination window not opening until Jan. 8. Moreover, the kind of change that Engine calls for here is generally not the kind of change that comes from an amicable settlement.

Ken Squire is the founder and president of 13D Monitor, an institutional research service on shareholder activism, and the founder and portfolio manager of the 13D Activist Fund, a mutual fund that invests in a portfolio of activist 13D investments. Viasat is owned in the fund.

Continue Reading

August 16, 2025

Datasets

We have compiled a comprehensive and well-structured dataset through the meticulous collection of image data obtained from the prestigious Institute of Radiotherapy and Nuclear Medicine (IRNUM) in Khyber Pakhtunkhwa, Pakistan. Utilizing the advanced and state-of-the-art GE SIGNA PET/MRI scanner, we acquired a comprehensive set of imaging data, comprising accurately captured and processed attenuation-corrected and reconstructed PET images, along with the invaluable gadolinium-enhanced T1-weighted images.

Our dataset encompasses a total of 207 meticulously selected digital imaging and communications in medicine (DICOM) images, forming the foundation for our rigorous image analysis endeavors. This intricate analysis process was meticulously carried out employing industry-standard tools such as MATLAB and the renowned imlook4d, ensuring the highest degree of precision and accuracy throughout the analytical pipeline.

To ensure a harmonious and coherent integration of the PET and T1-weighted images, a crucial preprocessing step involved the meticulous resampling of the PET matrix, precisely aligning it with the exact slice positions of the T1-weighted images. Subsequently, to facilitate standardized and consistent quantitative analysis, a vital step encompassed the normalization of voxel values to standardized uptake values (SUV), ensuring the establishment of a common metric for accurate assessment and comparison. This normalization process followed the formulation provided by Eq. (3), seamlessly integrating the essential mathematical framework into our analytical workflow.

While FLT and Gd have emerged as widely utilized PET tracer and MRI contrast agent, respectively, for the assessment of brain tumors, it is imperative to acknowledge their inherent limitations. Notably, FLT uptake, although serving as an indicator of cellular activity, lacks specificity when discerning malignant neoplasms, as it can be influenced by factors such as increased permeability resulting from blood-brain barrier disruption, which may also manifest in bone marrow and treatment-induced alterations. On the other hand, Gd, despite its established utility, necessitates careful chelation to ensure safe utilization in MRI due to its inherent toxicity. Moreover, differentiating between tumor tissues and surgical invasions poses a formidable challenge.

In this study, we employed 18 F-fluorothymidine (18 F-FLT) as the PET tracer for imaging glioblastoma multiforme (GBM). 18 F-FLT is a thymidine analog that functions as a proliferation-specific radiotracer by targeting thymidine kinase-1 (TK1), an enzyme upregulated during DNA synthesis in actively dividing cells. FLT is phosphorylated intracellularly and retained within proliferating cells, making it a valuable marker for assessing tumor cell proliferation. This characteristic renders 18 F-FLT a “tumor-specific” tracer, particularly advantageous for identifying active tumor regions beyond the contrast-enhancing zones visible on MRI. Unlike amino acid tracers such as 18 F-FET or 11 C-methionine, which accumulate based on increased transport in tumor cells, FLT provides more direct insight into mitotic activity, thereby offering complementary biological information that enhances the delineation of aggressive tumor subregions and contributes to more informed treatment planning and monitoring.

All imaging data were acquired using the GE SIGNA PET/MRI hybrid scanner, which integrates simultaneous PET and MRI acquisition. This system enabled exact co-registration between PET and MR images by capturing both modalities during a single imaging session without repositioning the patient. Consequently, issues related to alignment or time-lag variability between modalities were inherently minimized. The MRI component involved contrast-enhanced T1-weighted imaging, using gadolinium-based contrast agents to visualize the enhancing tumor core. Simultaneously, 18 F-FLT PET images were acquired to assess tumor proliferation. For PET imaging, attenuation correction was automatically performed using MRI-based correction algorithms native to the hybrid scanner. No separate image smoothing was applied during acquisition.

To evaluate surgical efficacy and determine the presence of any residual tumor tissue, the integration of follow-up MRI with contrast enhancement within a 48-hour timeframe following surgery assumes paramount importance. Such subsequent examinations often reveal contrast enhancement attributed to surgical intervention and the effects of radiotherapy. As shown in Fig. 1, pre-surgical FLT-PET/MRI brain imaging highlights glioblastoma contrast enhancement on Gd-enhanced MRI and FLT-PET activity concentration. The PET image was normalized to create a Gd-enhanced T1-weighted MRI matrix. While FLT PET and contrast-enhanced MRI images of GBM offer invaluable insights and information, it is crucial to consider and account for the aforementioned limitations. Therefore, exploring alternative imaging methodologies and approaches is imperative to enhance the accuracy of diagnosis and improve the efficacy of treatment monitoring, ultimately augmenting patient outcomes and prognoses.

Fig. 1

Pre-surgical FLT-PET/MRI brain imaging showing glioblastoma contrast enhancement on Gd-enhanced MRI and FLT-PET activity concentration. PET image was normalized to create a Gd-enhanced T1-weighted MRI matrix.

The normalization process followed the formulation provided by Eq. (3), which can be expressed as:

$$SUV=~frac{{{I_{PET}}}}{{{C_t}{text{*}}{A_{inj}}*W*D}}$$

Where:

(SUV) represents the Standardized Uptake Value. ({I_{PET}}) denotes the measured intensity in the PET image. ({C_t}) stands for the tissue concentration of the tracer. ({A_{inj}}) represents the injected activity concentration. W denotes the weight of the patient. D represents the quantity of tracer injected adjusted for decay.

In this equation, the measured PET image intensity represents the pixel intensity value obtained from the PET image. The tissue concentration of the tracer refers to the concentration of the radiotracer within the tissue of interest. The injected activity concentration denotes the concentration of the injected radiotracer dose. The patient’s weight is the weight of the individual undergoing the PET scan. Lastly, the injected fluid tracer accounts in the the decay-corrected for the radioactive decay of the tracer over time.

Digital image processing

ROI selection

In the realm of brain tumor imaging, the current study has made significant advancements in the methodologies employed for precise delineation of tumor regions within PET and MRI images through a series of preprocessing steps. Nonetheless, the task of accurately delineating regions of interest (ROIs) in PET images presents challenges due to the influence of the partial volume effect (PVE), which impacts the resolution of PET cameras and results in a low signal-to-noise ratio.

To mitigate these challenges, a meticulous approach was adopted wherein tumor regions were delineated on each transaxial slice of the PET and MRI scans. Notably, the MRI scans exhibited superior resolution and contrast properties, enabling enhanced visualization of tumor regions within the ROIs.

Partial Volume Effect (PVE) Correction Equation:

$$PV{E_{Correcte{d_{PET}}}}=frac{{PET_intensity~}}{{PVE_Factor}}$$

Where:

(PET_Intensity) represents the intensity measured in the PET image. (PVE_Factor) denotes the correction factor accounting for the partial volume effect.

In order to establish initial delineation of PET ROIs, two distinct methods were devised. The first method involved excluding meninges and skull bone when tumors were situated in proximity to these areas. Conversely, the second method entailed comparing the affected cerebral hemisphere with its contralateral counterpart in instances where regions of heightened radiotracer uptake were not located near the proximity to the skull bone or meninges. Subsequently, a delineated ROIs were subjected to adaptive thresholding techniques, refining the initial approximation. In the image analysis phase, a synergistic combination of both delineation methods was employed to improve the accuracy of tumor delineation and achieve robust ROIs for subsequent volume of interest statistics and standardized uptake value (SUV) analyses. This study underscores the significance of advancing tumor imaging techniques and highlights the potential advantages associated with utilizing a comprehensive amalgamation of methodologies to surmount limitations such as the PVE. By refining and optimizing these techniques, it becomes conceivable to enhance the precision and specificity of brain tumor diagnosis and treatment monitoring, thereby yielding improved patient outcomes.

Adaptive Thresholding Equation:

$$Threshold=Mean+k*Standard~Deviation$$

Where:

(Mean) represents the mean intensity within the delineated region. (Standard~Deviation) denotes the standard deviation of intensity within the delineated region. k represents a constant multiplier used to adjust the threshold.

To enhance the precision of tumor delineation in PET/MRI imaging, it is essential to refine the imaging process by excluding structures such as the skull bone and meninges. As shown in Fig. 2, this refinement involves copying the PET delineation onto the MRI image to define initial boundaries (Fig. 2a and b). Subsequently, the MRI is employed to exclude the skull bone and meninges from the region of interest (ROI) (Fig. 2c). The delineation is further enhanced using adaptive thresholding techniques (Fig. 2d). These refinements are critical for achieving accurate boundary definitions and improving diagnostic precision.

Contrast-enhanced MRI and FLT-PET imaging provide complementary insights into tumor characterization. As illustrated in Fig. 3, the blue areas in the contrast-enhanced MRI image (Fig. 3a) represent active tumor regions identified through increased uptake of the contrast agent. Meanwhile, the red delineated regions in the FLT-PET image (Fig. 3b) highlight tumors with high proliferative activity, offering a distinction from the surrounding tissues. This combined imaging approach enhances the accuracy of tumor detection and characterization, aiding in treatment planning and monitoring.

For tumor volume delineation, distinct segmentation strategies were applied to PET and MRI modalities. The MRI tumor volumes were manually segmented on contrast-enhanced T1-weighted images using the imlook4d analysis platform by two independent expert radiologists with over five years of neuroimaging experience. To reduce inter-observer variability, consensus segmentation was used for final volume generation. The inter-observer agreement was assessed using the Dice Similarity Coefficient (DSC), yielding an average Dice score of 0.88 ± 0.04 across subjects. For PET imaging, an adaptive thresholding approach was used to delineate the metabolic tumor volume (MTV). Specifically, regions with uptake values exceeding 40% of the lesion’s SUV_max were classified as tumor regions, in accordance with previously published clinical guidelines. This method has been shown to provide robust segmentation for 18 F-FLT PET imaging in glioma patients. The thresholding algorithm was implemented in MATLAB and validated internally through comparison with manually segmented test cases. To quantify the spatial agreement between PET- and MRI-derived tumor volumes, the Dice Similarity Coefficient was computed for each subject. The average Dice coefficient observed across all patient examinations was 0.42 ± 0.09, consistent with prior reports indicating limited spatial overlap between functional (PET) and structural (MRI) imaging in GBM. This reinforces the complementary nature of the two modalities and highlights the clinical relevance of multimodal imaging in glioblastoma assessment.

Consistency of inter observer

To assess the reliability and reproducibility of manual tumor delineation, a rigorous examination was conducted, involving a subset of four randomly selected patients from the overall cohort. In this evaluation, three proficient individuals expertly delineated regions of interest (ROIs) in a total of 20 MRI examinations, thereby yielding a comprehensive dataset of 60 MRI examinations for the purpose of consistency analysis. The degree of similarity between the segmented tumor volumes was quantitatively assessed by means of the Dice index, a widely recognized metric for measuring spatial overlap between binary segmentations.

Dice Index Equation:

$$Dice_Index=~frac{{2*left| {RO{I_1} cap RO{I_2}} right|}}{{left| {RO{I_1}} right|+left| {RO{I_2}} right|}}*100%$$

Where:

(RO{I_1}) and (RO{I_2}) represent two segmented tumor volume. (left| {RO{I_1} cap RO{I_2}} right|) denotes the intersection of the two segmented volumes. (left| {RO{I_1}} right|) and (left| {RO{I_2}} right|) represent the total volume of each segmented tumor. The Dice Index quantifies the extent of spatial intersection between the two segmented tumor volumes, represented in percentage terms.

In order to gain deeper insights into the consistency of the manual delineation process, the coefficient of variation (CV) was employed to calculate the relative standard deviation of the segmented tumor volumes. This statistical measure allowed for a comprehensive analysis of the degree to which the delineated tumor volumes deviated from the mean volume. By examining the mean tumor volumes associated with different delineated ROIs for each examination, the CV served as a valuable indicator of the degree of clustering or dispersion of the data points around the mean volume.

Coefficient of Variation (CV) Equation:

$$CV=frac{{Standard~Deviation}}{{Mean~Volume}}*100%$$

Where:

(Standard~Deviation) represents the standard deviation of the segmented tumor volumes. (Mean~Volume) denotes volume of the tumor being segmented with highlighted mean. The (CV) calculates the relative standard deviation of the segmented tumor volume as a percenatage of the mean volume.

This meticulous evaluation shed light on the subjective nature of manual segmentation in tumor delineation. It became evident that the manual approach introduced a certain level of variability, highlighting the need for more objective and standardized methods in the realm of tumor delineation. The findings underscored the importance of adopting quantitative and reproducible techniques to enhance the accuracy and reliability of tumor delineation processes, ultimately contributing to improved diagnostic and treatment outcomes.

Patient cohort and study design

This was a prospective observational study conducted at the Institute of Radiotherapy and Nuclear Medicine (IRNUM), Khyber Pakhtunkhwa, Pakistan, with institutional ethical approval obtained prior to data collection. All patients provided written informed consent in accordance with the Declaration of Helsinki. The study enrolled 22 patients diagnosed with histologically confirmed glioblastoma multiforme (GBM), who were undergoing initial staging and treatment planning. Inclusion criteria comprised: (1) newly diagnosed and treatment-naïve GBM patients; (2) availability of both pre-treatment PET and MRI scans; and (3) no prior radiotherapy, chemotherapy, or neurosurgical intervention except for biopsy. Patients were excluded if they had incomplete imaging data, motion artifacts, or were lost to follow-up. Ultimately, a total of 18 patients met all inclusion criteria and were analyzed. All imaging was conducted within a narrow time window, with PET and MRI scans performed within 48 h of each other to minimize temporal variations in tumor volume. Both imaging modalities were acquired before initiation of any radiochemotherapy or surgical resection to ensure a consistent baseline across the cohort. This approach eliminated potential confounding factors related to treatment-induced changes in enhancement or uptake patterns. Potential biases were mitigated by adopting strict inclusion criteria and standardized imaging protocols. However, we acknowledge that selection bias may persist, as patients undergoing both PET and MRI are often those with more complex or ambiguous presentations. This limitation is discussed in the concluding section. Nevertheless, the consistency in disease stage and imaging timelines across the included patients enhances the reliability of volume comparisons between modalities.

Segmentation performance evaluation

The inherent limitations of imaging systems, particularly their restricted spatial resolution, give rise to partial volume effects (PVEs) that engender spill-out phenomena in small objects or regions, as vividly depicted in Fig. 4. Consequently, when delineating ROIs for analysis, the pervasive influence of PVEs must be duly taken into account. In order to establish a reliable reference point for accurate measurements, a ground truth image was meticulously constructed by amalgamating PET images with predetermined standardized uptake values (SUVs) corresponding to distinct anatomical components, such as the tumor, regions with elevated tumor activity, skull bone, and adjacent background areas. These ground truth images serve as invaluable benchmarks, enabling precise quantification of object dimensions within the imaging domain.

Figure 4 serves as a visual illustration, juxtaposing a phantom image on the left and a conventional PET image on the right. Evidently, the PET image exemplifies a discernible attenuation of activity in diminutive entities, such as the high tumor spots region, which regrettably fails to manifest in the phantom image as prominently indicated by the absence of its characteristic red coloration. Furthermore, a notable discrepancy in spatial resolution is discernible between these two images, underscoring the criticality of addressing PVEs within imaging systems to foster enhanced precision in ROI calculations and tumor delineation endeavors.

By actively mitigating the deleterious effects of PVEs through judicious methodological interventions, one can effectively ameliorate the impact of spatial resolution constraints, thereby bolstering the accuracy and reliability of ROI determinations. This imperative pursuit toward PVE-aware imaging practices holds considerable promise in refining tumor delineation methodologies, ultimately facilitating more robust and dependable analyses with significant ramifications for diagnostic and therapeutic decision-making.

The impact of partial volume effects (PVE) on imaging accuracy can be observed in a constructed phantom study. As shown in Fig. 4, the phantom with precise dimensions is depicted on the left, while the original PET image on the right demonstrates blurred object boundaries and a loss of activity in small objects (indicated by the white arrow) due to PVE. This highlights the need for advanced imaging techniques to mitigate PVE and improve resolution and activity quantification in PET imaging.

To appraise the efficacy and discriminatory capabilities of diverse segmentation algorithms, a comprehensive assessment was conducted utilizing a meticulously constructed phantom and the generation of two synthetic PET images that faithfully emulated the salient characteristics of a genuine PET image dataset. With a meticulous attention to detail, these synthetic images were meticulously engineered to exhibit high and low tumor-to-background ratios (TBR), while maintaining a pixel size of 0.4883 mm in both the x- and y-directions, along with a slice spacing in the z-direction up to 1 mm. Notably, the “Uptake” values for FLT corresponding to each TBR configuration were meticulously determined and tabulated for reference in Table 1, serving as invaluable benchmarks for subsequent analyses¹.

To faithfully replicate the inherent blurriness that frequently plagues real-world PET images, a Gaussian smoothing filter was judiciously employed to deliberately introduce blur effects onto the synthetic images. Specifically, the input parameters for the full width at half maximum (FWHM) of the Gaussian smoothing filter were diligently set at 3.5 mm in both the x- and y-directions, with a corresponding value of 5 mm in the z-direction. Such deliberate manipulation effectively induced blurred boundaries, faithfully mirroring the commonplace phenomenon encountered in actual PET images¹.

By subjecting these meticulously crafted synthetic PET images to a comprehensive evaluation, profound insights were gleaned regarding the performance characteristics of the segmentation algorithms under scrutiny. Specifically, the algorithms’ capacity to accurately discern tumor activity from background activity was scrutinized, shedding light on their respective strengths and limitations¹. This rigorous evaluation process constitutes an invaluable contribution to the field, offering crucial guidance and empirical evidence to guide the selection and refinement of segmentation algorithms in pursuit of heightened accuracy and reliability in tumor delineation endeavors.

Table 1 FLT uptake values (in kBq/mL) for different regions used to create synthetic PET images with varying tumor-to-background ratios.

Within the confines of Table 1, an intricate compilation of four distinct anatomical regions is presented, accompanied by their respective FLT uptake values crucial for the creation of PET images boasting divergent tumor-to-background ratios. With meticulous attention to detail, these FLT uptake values are meticulously expressed in kilobecquerels per milliliter (kBq/mL), precisely reflecting the standardized units inherent to PET imaging. It is noteworthy that the high uptake tumor spots, in particular, manifest significantly elevated FLT uptake values when juxtaposed with the remaining regions, possibly indicative of a more pronounced presence of metabolically active tumor tissue within these localized areas. Conversely, the values assigned to the skull bone and background regions exhibit an identical manifestation in both the high and low FLT uptake images, effectively indicating a relative paucity in terms of activity levels within these anatomical regions.

To comprehensively evaluate the segmentation accuracy of diverse thresholding methodologies, a meticulous analysis was conducted employing two distinct synthetic PET images thoughtfully engineered to encapsulate disparate tumor-to-background ratios, as comprehensively expounded upon in Sect. 3.2.4. In order to gauge the precision of the segmentation techniques under scrutiny, the Dice index, renowned for its efficacy in quantifying the degree of similarity between the segmented volume and the ground truth volume, was deftly harnessed for evaluation purposes. In accordance with established conventions, this Dice index was diligently calculated by multiplying twice the cardinality of the common elements shared by both the segmented and ground truth volumes, subsequently dividing this product by the element’s total cardinalities of the within respective group. Mathematically articulated the Dice (V1, V2) = 2 × (|V1 ∩ V2″https://www.nature.com/”V1| + |V2|)¹⁷, this esteemed metric definitively enabled the precise quantification of the segmentation accuracy achieved by each thresholding methodology under scrutiny, thereby facilitating an informed assessment of their relative efficacy.

Methods of PET volume segmentation

Medical Image segmentation, a fundamental aspect of medical image processing, assumes paramount significance in facilitating subsequent computational analysis. Its underlying objective entails partitioning images into distinct regions or segments to enable efficient handling of individual components. Among the manifold techniques employed for this purpose, thresholding emerges as a prominent approach, particularly adept at converting grayscale images into binary counterparts. This transformative process involves assigning foreground or background values to pixels whose intensity surpasses or equals a predetermined threshold value, thereby generating a binary mask that selectively designates pixels of interest as 1 while relegating others to a value of 0.

Within the scope of the present research endeavor, the paramount focus revolves around accurately delineating FLT-PET target cross-sections and effectively characterizing tumor volumes. To achieve this crucial objective, a comprehensive evaluation encompassing three distinct thresholding segmentation methodologies was undertaken, meticulously tailored to the specific nuances of the PET image dataset at hand. Notably, this discerning analysis incorporated two conventional thresholding techniques, esteemed for their established efficacy, alongside an adaptive thresholding technique, celebrated for its capacity to dynamically adapt to the unique characteristics inherent to the dataset under scrutiny. By judiciously applying these diverse segmentation approaches, the research aimed to achieve precise and robust tumor volume definition, thus paving the way for subsequent comprehensive analysis and interpretation of the FLT-PET image data.

The conventional thresholding methods conventionally employed a uniform threshold value applied to all pixels within the image, assuming a homogeneous response across the entire image domain. However, the efficacy of this approach is contingent upon various image attributes such as texture, noise characteristics, and the employed image reconstruction techniques. Recognizing the inherent limitations of a global thresholding strategy, an innovative and advanced adaptive thresholding technique was incorporated, wherein distinct threshold values were assigned to individual pixels based on their specific image properties.

This adaptive thresholding methodology, distinguished by its intrinsic adaptability to account for spatial variations in image illumination, proved to be remarkably robust and well-suited for the PET dataset under investigation. Unlike fixed threshold approaches, which faltered in the face of inherent spatial differences in image illumination, the adaptive thresholding method adeptly responded to such variances, facilitating accurate and refined segmentation outcomes. Notably, the adaptive thresholding algorithm was applied to the initial rough delineation of the regions of interest (ROIs), as visually depicted in Fig. 5.

In order to ascertain the fidelity and precision of the segmentation achieved through the diverse segmentation methods, a pivotal metric known as the Dice index was leveraged. This quantitative index, extensively employed in the field of image segmentation, facilitated a comprehensive evaluation of the segmentation accuracy vis-à-vis the ground truth-constructed tumor volume. By quantifying the overlap and concordance between the segmented regions and the true tumor volume, the Dice index served as an indispensable tool in assessing the reliability and robustness of the employed image segmentation techniques.

The implementation of adaptive thresholding, an intricate technique rooted in the concept of local mean intensity surrounding each pixel, was realized in MATLAB through the utilization of the integral image method. A critical aspect of this approach involved determining the optimal sensitivity factor, which played a pivotal role in delineating the pixels deemed to belong to the foreground. To determine the most suitable sensitivity factor, an exhaustive evaluation of the PET imaging dataset was undertaken, employing two distinct PET images, each encompassing regions of varying TBR. Notably, these PET images were accompanied by preliminary delineations of the regions of interest (ROIs) to guide the analysis.

Within this evaluation framework, a comprehensive range of sensitivity factors spanning the spectrum of 0.1 to 0.9 was systematically explored. The accuracy and fidelity of tumor boundaries were meticulously assessed using a robust metric known as the Dice index. Through the rigorous analysis, the adaptive thresholding method emerged as the paradigmatic choice, yielding the most precise and reliable segmentation outcomes for the lesions under scrutiny.

It is worth emphasizing that the accurate definition of the ROIs assumes paramount significance, as even minute errors or inaccuracies within the delineated margins can impart substantial repercussions on crucial metrics such as SUV and TLA-based metrics. Thus, with utmost meticulousness, the SUV parameters were diligently calculated employing the well-established Eq. (3). Furthermore, the computation of the TLA further enriched the analytical repertoire, affording a more profound and nuanced comprehension of the tumor’s intricate characteristics.

These noteworthy findings not only contribute valuable insights to the existing body of knowledge pertaining to PET imaging data and tumor volume estimation through image segmentation but also hold substantial promise in informing and shaping future endeavors within this domain. Segmentation accuracy plays a critical role in the precise delineation of tumor boundaries in imaging studies. As demonstrated in Fig. 5, a transaxial slice comparison highlights the difference between traditional and adaptive thresholding methods. The adaptive approach shows improved accuracy in defining tumor regions, underscoring its potential for enhancing imaging precision and clinical decision-making.

Image features extraction

Sophisticated algorithms were meticulously crafted to extract intricate features from the PET and MRI image datasets, facilitating a comprehensive analysis of the distinctive characteristics and commonalities in tumor features. A central focus of this investigation was to evaluate the degree of tumor volume overlap (referred to as Voverlap) observed in both modalities. Precisely quantifying Voverlap entailed assessing the segmented volume shared by the VPET and VMR images. Furthermore, the analysis encompassed discerning the distinct volumes unique to each modality, denoted as VonlyPET and VonlyMR, respectively. It is crucial to comprehend that PET imaging enables the acquisition of information at an earlier stage in tumor development compared to MRI. Thus, VonlyPET signifies the tumor exhibiting active growth discernible solely through PET imaging, not yet manifest in MRI. Similarly, VonlyMR represents the active and necrotic tumor volume exclusively visualized through MRI imaging. By judiciously amalgamating the segmented volumes attributed to either VPET or VMR or both, their volume of tumor in combination, portrayed in MR and PET images can be ascertained. This comprehensive information significantly contributes to a more profound understanding of tumor growth dynamics and provides crucial insights into potential avenues for treatment optimization. The segmentation of tumor volumes across PET and MR modalities provides valuable insights into spatial overlap and modality-specific differences. As illustrated in Fig. 6, the segmented volumes for PET and MR are displayed, highlighting their spatial overlap (V_overlap) as well as unique volumes identified exclusively by each modality (V_onlyPET and V_onlyMR). This comparison underscores the complementary nature of PET and MR imaging for comprehensive tumor characterization.

Figure 7 exhibits a meticulously selected axial slice of the brain, ingeniously captured utilizing MRI with contrast enhancement. This striking visualization distinctly showcases the derived tumor volumes, meticulously delineating the distinct regions exclusively revealed through PET imaging (referred to as PET-only), the unique areas solely visible in MR imaging (referred to as MR-only), and the captivating overlapping tumor volumes that are prominently observed in both modalities. This comprehensive illustration provides a profound visual representation of the intricate interplay between PET and MR imaging in capturing and characterizing tumor volumes, thereby enriching our understanding of the multifaceted nature of these tumors.

Statistical analysis

All statistical analyses were performed using MATLAB and SPSS. To quantitatively compare tumor volumes derived from PET and MRI, we applied both descriptive and inferential statistical tests. A paired t-test was used to evaluate whether statistically significant differences existed between PET-derived and MRI-derived tumor volumes across matched examinations. For datasets violating normality assumptions (as verified using the Shapiro-Wilk test), the Wilcoxon signed-rank test was employed as a non-parametric alternative. A two-tailed significance level of p < 0.05 was considered statistically significant. The relationship between PET and MRI tumor volumes was assessed using Pearson’s correlation coefficient (r). In cases of non-normal data distribution, Spearman’s rank correlation (ρ) was used. However, we acknowledge that correlation reflects association and not agreement.

To assess the level of agreement and identify any systematic bias between PET and MRI tumor volumes, we conducted a Bland-Altman analysis. Bland-Altman plots were generated to illustrate the mean difference (bias) between modalities and the 95% limits of agreement, defined as the mean difference ± 1.96 standard deviations. This approach provided a visual and statistical measure of proportional bias and potential outliers. For each comparison, 95% confidence intervals were reported alongside p-values to support the interpretation of effect sizes and measurement variability.

Given the limited cohort size, we treated statistical outcomes as preliminary. We explicitly note the limited statistical power and advocate for further validation using larger patient cohorts in future studies. As shown in Fig. 8, the Bland-Altman analysis illustrates a mean bias of 3.0 cm³, indicating that PET tends to slightly overestimate tumor volume compared to MRI. As summarized in Table 2, statistical analysis revealed a significant difference between PET- and MRI-derived tumor volumes, with PET consistently measuring larger volumes.

Table 2 Summary of statistical comparison between PET- and MRI-derived tumor volumes.

Imaging protocol and PET tracer details

All imaging was conducted using a hybrid GE SIGNA PET/MRI scanner at the Institute of Radiotherapy and Nuclear Medicine (IRNUM), enabling simultaneous acquisition of PET and MRI data to eliminate registration errors and temporal discrepancies. PET and MRI scans were performed on the same day during each of the four scheduled examination sessions.

We employed the radiotracer 3’–deoxy-3’–[^18F]fluorothymidine (FLT), an ^18F-labeled thymidine analog, as the PET tracer. FLT is a tumor-specific proliferation marker that accumulates in actively dividing cells by mimicking endogenous thymidine, a DNA synthesis substrate. Due to its uptake via thymidine kinase-1 (TK1), which is upregulated in proliferating glioma cells, FLT-PET provides a reliable assessment of tumor growth and cellular activity. For the PET acquisition, approximately 370 MBq (10 mCi) of ^18F-FLT was intravenously administered, followed by a 60-minute uptake period prior to scanning. PET data were acquired in 3D mode for 20 min and corrected for attenuation using MRI-based attenuation maps. MRI scans included a T1-weighted post-contrast sequence (Gd-enhanced) with the following parameters: repetition time (TR) = 500 ms, echo time (TE) = 15 ms, slice thickness = 1 mm, field of view (FOV) = 256 × 256 mm. The contrast agent used was gadolinium-DTPA, administered at a dose of 0.1 mmol/kg body weight immediately prior to scanning. All PET and MRI images were preprocessed using MATLAB and imlook4d, including intensity normalization, motion correction, and resampling to a common voxel size. For PET-MRI co-registration, the hybrid system enabled automatic alignment, and further manual adjustments were made when necessary to ensure spatial fidelity.

Category: 3. Business

Battery sustainability index (BSI)

PV power behavior

Load behavior

Holistic sustainability index

Economic assessment of PV-battery systems

Levelized cost of energy (LCOE)

The battery replacement time

South Korea’s vital statistics

Commensal intestinal bacteria can protect shrimp from the pathogenic Vibrio infection

Commensal bacteria and phage can supress Vibrio pathogen growth

Timing of commensal and phage administration is critial for effective colonization resistance

Commensal bacteria supress pathogen by nutrient competition and prophage induction

Synergistic interaction of commensal microbes and phage confers colonization resistance against pathogenic Vibrio in shrimp

AI Welfare: When the Model Walks Away

A Measured Advance in AI Safety

Critics Raise Important Questions

The Bigger Picture in AI Regulation

Features and Equipment

Pricing and Payment Schedule

Availability

Company: Avantor (AVTR)

Activist: Engine Capital

What’s happening

Behind the scenes

Datasets

Digital image processing

ROI selection

Consistency of inter observer

Patient cohort and study design

Segmentation performance evaluation

Methods of PET volume segmentation

Image features extraction

Statistical analysis

Imaging protocol and PET tracer details