Blog

  • Last-gasp shot secures Iceland quarter-finals berth, Czechia enhance chances with big win over Japan

    Spain led by one goal with 20 seconds left in a crucial match against Iceland, but conceded two goals in the last 16 seconds to concede a 32:31 loss against the Nordic side. Both teams qualified to the quarter-finals, at Serbia’s expense.

    In Group IV of the main round, Czechia needed to erase a 13-goal deficit against Egypt and they did their part, securing a 42:27 against Japan, forcing the hosts to take at least a point against Denmark to qualify for the quarter-finals.

    Main round
    Group II 
    Spain vs Iceland 31:32 (15:15)  

    Spain struggled in the opening minutes, unable to find clear shooting positions. When Jens Sigurdarson added two more saves to his tally, Iceland capitalised to break away and open a four-goal lead. It was clear Spain needed a tactical shift, which arrived after a time‑out.  

    Switching to a 5–1 defence, Spain began to disrupt Iceland’s rhythm, successfully isolating their centre back. That defensive change paved the way for a comeback, and midway through the first half Sergio Sánchez Vidán scored to make it 8:9.  

    Spain kept pressing, looking to push Iceland to a breaking point. When they finally levelled at 11:11, the match was back on even terms. With both sides showing 54% attacking efficiency, the statistics told the story of a first half that was evenly matched — Spain growing stronger towards the break, and Iceland losing a touch of focus.  

    Spain’s momentum carried into the second half, as they opened with a 2:0 run to claim a 17:15 lead — their first of the match. From there, it became a nip‑and‑tuck battle, with both sides briefly holding narrow leads, underlining the high stakes.  

    The drama continued right to the end. Goalkeeper Sigurjón Bragi Atlason stood tall for Iceland when they needed him most, turning a 24:25 deficit into a 28:27 lead. But nothing could prepare the crowd for the final twists.  

    The title holders stole possession twice in quick succession, punishing Iceland for playing without a goalkeeper and snatching the lead at 31:30. Iceland levelled again, and then one costly Spanish mistake — a missed shot on an empty goal — left the door open. With just seven seconds remaining, Ágúst Gudmundsson struck for his ninth goal of the match, almost on the buzzer, sealing two vital points for Iceland.  

    The result pushed both teams into the quarter-finals, eliminating Serbia before their final group game against Saudi Arabia.  

    Player of the Match: Ágúst Gudmundsson (Iceland)  

    Group IV
    Japan vs Czechia 27:42 (12:19)

    Missing out on a win against Egypt was always going to be costly for Czechia, which had a three-goal lead with three minutes left on the clock in the main round opener. But the European side still had a fighting chance.

    First, they needed to win against Japan. Then, they needed Egypt to lose against Denmark. And in the process, a 13-goal difference had to be wiped out, as the African champions enjoyed bigger wins against their opponents.

    Easier said than done for Czechia, but they surely threw off  the match knowing exactly what they needed, as they were leading 11:4 after 15 minutes. Sure, Japan were in no man’s land after their previous losses, but they still produced some high-quality handball throughout the competition.

    And Czechia continued to pounce, continously finding new sollutions in attack, until they hit a wall and failed to score for five minutes and 21 seconds. A Japan 3:0 run cut the gap to only four goals, 14:10, but at the end of the first half, Czechia were still running away with the win, leading 19:12, and applying some pressure to Egypt.

    There was still a mountain to be climbed, as at that point the gap was only six goals to Egypt, which were slated to end the day against Denmark, therefore every possible goal was important for the European side, which missed some crucial chances to open the first eight-goal lead between the 34th and the 36th minutes.

    A 7:2 run, though, fueled by two goals from Daniel Mahušek and Ondřej Šíma, helped Czechia exert even more pressure in their duel with Egypt, as they opened up a 11-goal lead, 31:20, with 12 minutes to go, playing their part excellently in order to boost their chances of progression to the quarter-finals.

    Not only Czechia did cancel Egypt’s advantage in goal difference, they did it in style, with seven of their goals coming on fast breaks, while probably delivering their best attacking performance of the competition, finishing the match with a 70% attacking efficiency, while goalkeeper Kryštof Pleva had 15 saves, for a 38% saving efficiency.

    The 42:27 Czechia win means that the European side improved drastically their chances of progression, needing only Denmark to win against Egypt, to secure their maiden quarter-finals appearance at the IHF Men’s Youth World Championship.

    Japan’s third consecutive loss sees the Asian side now heading to the Placement Matches 13-16, where they are due to face Saudi Arabia.

    Player of the Match: Ondřej Šíma (Czechia)

    President’s Cup
    Group III
    Faroe Islands vs Argentina 26:26 (15:14)

    The Faroe Islands made a strong start against Argentina in their bid for a second President’s Cup victory, and for almost 20 minutes, it looked well within reach. In that spell, the Faroese led by as many as four goals (12:8), using a 7‑on‑6 attack to maintain their advantage.  

    However, late saves from Simon Viera and a series of rushed Faroese attacks stalled what had been an impressive half. Argentina found fresh energy, stepped up a gear, and in the closing stages of the first period, narrowed the gap to just one goal — setting up a fascinating second half.  

    Argentina levelled early after the restart and, driven by their top scorer Nicolás La Delfa, claimed their first lead since the opening minutes. The Faroe Islands lacked the sharpness to wrest control back, while Argentina pressed hard for what would have been a notable win at Egypt 2025.  

    With ten minutes remaining, the teams were locked at 21:21, but this time it was the Faroe Islands chasing as Argentina edged closer to victory. Yet in the very last second, Rúnar Hammer stunned the South Americans with a decisive strike to secure a draw.  

    The Faroe Islands will continue in the Placement Matches 17–20, while Argentina move on to the Placement Matches 21–24.  

    Player of the Match: Rúnar Hammer (Faroe Islands)

    Group IV
    Bahrain vs Tunisia 29:29 (15:16)

    Thanks to Mohamed Abdulhusan’s string of saves in the closing minutes, Bahrain salvaged a draw against Tunisia. However, it was the African side who topped the group on goal difference and will continue fighting for the President’s Cup in the Placement Matches 17–20, while Bahrain will proceed to the Placement Matches 21–24.

    While his Tunisian colleague Mohamed Chabchoub made 11 saves and was named Player of the Match, Abdulhusan had a big impact in the last quarter of the game after coming on as a substitute. He saved six of eight shots for a 75% save rate, denying the African side a single goal in the last 10 minutes.

    The opening half featured a very tight contest, as neither team could pull clear by more than two goals. Bahrain were the better side in the opening minutes, taking a 4:2 lead, but Tunisia held a 14:12 advantage nearly six minutes before the break.

    Two goals from Ali Aryash then launched Bahrain’s 3:0 run, handing them a 14:14 lead. But the African side netted twice in the closing minute of the half, heading to the dressing room with a one‑goal cushion.

    The pattern of the game did not change after the restart, as both rivals took turns in the lead but were unable to create even a two‑goal gap for a long time. When Bahrain missed a few shots — including one from the seven‑metre line — Tunisia gained momentum, and a strong 5:0 run saw them take a 28:23 lead with 12 minutes remaining.

    But the fight was not over. Abdulhusan’s vital saves, combined with Salman Alshowaikh’s goals, helped Bahrain draw level in the 58th minute. However, neither team could score in the remaining time, and they ultimately shared the spoils in this thriller.

    Player of the Match: Mohamed Chabchoub (Tunisia)

    Continue Reading

  • Michelin-Starred Chef Giorgio Locatelli on His New Caravaggio-Inspired Cuisine

    Michelin-Starred Chef Giorgio Locatelli on His New Caravaggio-Inspired Cuisine

    Art

    Maxwell Rabb

    Portrait of Giorgio Locatelli in front of Caravaggio’s Supper at Emmaus (ca. 1601). Photo by Lisa Linder. Courtesy of the National Gallery.

    For the Michelin-starred chef Giorgio Locatelli, the most important thing a restaurant can serve isn’t a dish or a perfect wine pairing but the feeling of eating among others. “The sharing of food—the sense of conviviality—is something that makes us feel better,” he told Artsy recently over the phone.

    A similar sentiment could apply to the collective ritual of viewing art at museums: people gathering in one place, drawn by beauty and history. At his new eponymous restaurant at London’s eminent National Gallery, Locatelli hopes to provide guests an opportunity to experience “quietness and relaxation” after exploring the museum’s collection, which includes the likes of Paul Cézanne through to Canaletto and J.M.W. Turner. With his menu, Locatelli hopes to complement the visual stimuli. “The flavor is what is needed here, just to equalize, as I say, the stimulation that you get in the gallery, which is very visual,” he said. “I want to do that with the taste.”

    Interior view of Locatelli at the National Gallery. Photo by Lisa Linder. Courtesy of the National Gallery.

    Doing things with taste is something Locatelli knows a thing or two about. Acclaimed for his refined cooking that blends traditional recipes with contemporary presentation, Locatelli is among the most celebrated Italian chefs working today. His roots in London’s food scene run deep. He was awarded Michelin stars for his two restaurants in the city: Locanda Locatelli, opened in 2002, and Zafferano, opened in 1999. He’s also a frequent presence on television and bestseller lists, and now he’s taking his renowned gastronomy into one of London’s most hallowed institutions.

    The 80-seat restaurant—which is accompanied by an espresso and maritozzi bar named Bar Giorgio—is situated in the newly restored Sainsbury Wing of the gallery, which houses masterworks like Sandro Boticelli’s Venus and Mars (ca. 1485) and Matteo di Giovanni’s The Assumption of the Virgin (ca. 1474). But despite the lofty surroundings, Locatelli’s aims to “lighten up a little bit” with a menu he describes as “democratic and commercial.” Critics seem to agree: “Come for the The Entombment by Michelangelo […] and stay for the orecchiette with nettle pesto on the first-floor mezzanine next to the gift shop,” wrote The Guardian’s Grace Dent in a recent review.

    The restaurant’s launch prompts the notion of the chef as artist: After all, chefs combine creativity, technique, and sensory expression to create something that engages people emotionally and aesthetically. One could argue that instead of canvas and clay, their medium is food…

    Locatelli gently pushes back against the characterization. “I never felt like an artist. I felt like an artisan—somebody that can express his emotion through the dishes, or his experiences of life through the dishes,” he clarified. Chefs are too pressed for time, too pressured by costs, and too restricted by delivery to access the same type of creativity as an artist, he says. But art does play a core role in his approach to the restaurant.

    Take Michelangelo Merisi da Caravaggio’s The Supper at Emmaus (1601), one of the leading lights in the National Gallery’s collection. The colossal masterpiece depicts a risen Jesus Christ seated at a simple wooden table with two disciples, captured the instant they recognize him. Harsh light sharpens the scene—their startled gestures, the bread and fruit, and the jutting basket—while Christ’s open hand and the warm, domestic setting give the moment a striking sense of invitation. “The fact that you don’t need to be cultured, you don’t need to be rich, you don’t need to be something in order to see God,” Locatelli observed. The basket teeters at the table’s edge, he notes, “so you are invited to be part of this scene.”

    The restaurant serves a loaf of bread whole, a subtle nod to the loaf at the edge of the table in Supper at Emmaus. The intention is for it to be broken and passed around the table. “It’s one piece of bread…you open it up and give a piece to everybody who’s having dinner with you,” he said.

    This September, Locatelli will go a step further and debut a menu inspired by Caravaggio in commemoration of the artist’s 454th birthday.

    Interior view of Locatelli at the National Gallery. Photo by Lisa Linder. Courtesy of the National Gallery.

    Servings will be drawn directly from the painter’s life and imagery. The tasting menu opens with an artichoke dish in a nod to a notorious incident in which Caravaggio smashed a plate into a waiter’s face after a quarrel over how the vegetable was prepared. The main course references the still-life elements embedded in Caravaggio’s paintings, particularly the chicken on the table of Supper at Emmaus. Locatelli’s version features guinea fowl with grapes, a combination drawn from the textures and colors of the fruit and meat in the dramatic painting. Dessert is still taking shape, but Locatelli envisions it as the most theatrical plate of the evening, “bloody” in color, with chiaroscuro-like shadows worked into the presentation. “That was the painting that really inspired me, because it was an expression of the moment,” he said.

    Beyond Caravaggio, art is a continual source of inspiration and engagement for Locatelli. “A lot of artists really have great respect for chefs because the artist wants to produce an emotion when he makes his art,” Locatelli said. “Food is part of life. Life is part of art. Art is part of food and so on.”

    Portrait of Giorgio Locatelli. Photo by Lisa Linder. Courtesy of the National Gallery.

    While he jokes that “I don’t even have enough space in my house,” Locatelli notes that he owns a spinning painting by Damien Hirst, gifted to him after he cooked a private dinner for the artist. He names British artists Gary Hume and Rebecca Warren, as well as contemporary Italian sculptor Jago, among the artists he admires most. In Naples, Locatelli met Jago, who is known for realistic marble sculptures, and he left a particular impression. “To see this big piece of marble turning into sculpture was one of the most incredible experiences in my life,” Locatelli said.

    Locatelli often draws on art for ideas: Joan Miró’s “practicality” and dynamic compositions sparked new ideas for plating dishes, and John Constable’s layered greens, which are delicately depicted in his famed landscapes, once inspired a salad. “There was such a beauty on this green coming out,” he recalled, saying that he immediately “really tried to get by using natural color.”

    Photo by Lisa Linder. Courtesy of the National Gallery.

    Ingredients, for the chef, are themselves works of art: “The greatest artist overall is nature. Nature is the greatest artist that we can ever imagine.” For Locatelli, food and art are meant to be shared experiences. Just as the basket in Caravaggio’s scene seems to extend toward the viewer, inviting them in, a museum invites people to stand side by side before a work; a table invites them to sit together over a meal.

    “[We] try to really celebrate this great sense of conviviality, which is something that puts us on a different level than every other mammal, where the biggest guys eat first, the other ones eat after,” Locatelli said. “As humans, we have been sharing our food.”

    MR

    MR

    Maxwell Rabb

    Maxwell Rabb is Artsy’s Staff Writer.

    Continue Reading

  • Why trusting the process is my thing

    Why trusting the process is my thing

    Her U.S. Classic to U.S. Championships process has been out in the open.

    But Rivera has also made changes behind the scenes, one of the most notable coming on the vault.

    After struggling with keeping her head in while she twisted – something that can make a gymnast get “caught up” in the air and stop spinning – Rivera took a unique step back, opting to train Yurchenko one-and-a-halves instead of her typical double twist.

    Because of the different amount of rotation required for the two vaults (the one-and-a-half requires more because of its forward landing), gymnasts who compete double twists often only train full-twisting variations, not the one-and-a-half.

    But for Rivera, something clicked.

    “I know it sounds kind of crazy, but it helps my mind a lot,” she said. “I wasn’t blocking really, it was kind of mental. But right when I started going through the one-and-a-half, I kind of flipped the switch.”

    Her turn in Paris has flipped another switch: her confidence level.

    Rivera says even just watching Biles, Lee, Jordan Chiles and Jade Carey in practice taught her important lessons.

    “I still talk to them, and they give me really good advice,” she says of her Paris teammates. “They just tell me to do myself, do what I do in the gym and believe in myself.

    “The way that they practised helped me, too,” Rivera continued. “They look so confident, and they’re like, ‘Oh, I know I’m gonna hit,’ so it kind of gave me that confidence as well. It’s really been helping me this year.”

    Her journey is far from over – even this year, as she prepares for October’s World Championships in Jakarta. But one thing’s for sure: Rivera’s faith and belief in her process will continue to carry her forward as she leads Team USA into the future.

    Continue Reading

  • Exploring the Association of Antimicrobial Use with Serratia marcescen

    Exploring the Association of Antimicrobial Use with Serratia marcescen

    Introduction

    Serratia marcescens a facultative anaerobic Gram-negative bacillus belonging to the Serratia genus within the Enterobacteriaceae family, is a significant opportunistic pathogen in hospitals. Widely distributed in hospital environments, S. marcescens is hard to fully eradicate via physical or chemical disinfection. Its high virulence enables it to cause respiratory, urinary tract, bloodstream infections, meningitis, and surgical site infections, especially in patients undergoing invasive procedures, surgeries, with weakened immunity, or experiencing trauma.1 It spreads rapidly among hospitalized patients and has triggered numerous nosocomial outbreaks.2,3 According to CHINET data,4,5 the detection rate of S. marcescens increased from 0.99% (2483 strains) of all clinically isolated strains in 2019 to 1.15% (5120 strains) in 2023, marking a 16.16% rise. In recent years, the emergence of multidrug-resistant S. marcescens, including those resistant to carbapenems,6,7 has made detecting and controlling its spread a tough challenge.

    Currently, international data on the evolution of resistance in S. marcescens relative to antimicrobial exposure are limited, and this is also a relatively unexplored area in the research of S. marcescens. Antimicrobial selective pressure is a key driver of bacterial resistance. Our research endeavors to innovatively explore the macro-level quantitative relationship between the two. Theoretically, antimicrobial use precedes bacterial resistance.8,9 This study will explore the correlation between S. marcescens resistance rates and antimicrobial Defined Daily Doses (DDDs) from the previous quarter, aiming to inform strategies for curbing S. marcescens resistance and optimizing antimicrobial use in clinical practice.

    Materials and Methods

    Strain Source

    S. marcescens strains isolated from patients (2021–2024) were collected under strict aseptic conditions, including sputum, throat swabs, bronchoalveolar lavage fluid, blood, urine, pleural effusion, peritoneal effusion, and drainage fluid, and cultured according to the National Clinical Laboratory Procedures (4th edition).10 For identical strains isolated multiple times from a single patient sample, only the initial susceptibility result per drug was retained. Colonizing or contaminating strains were excluded to avoid confounding true pathogens, based on at least two of the following clinical-microbiological discordance criteria: (1) Absence of infection-related symptoms (eg, fever, purulent discharge, leukocytosis, or imaging findings); (2) Clinical improvement despite in vitro resistance to empiric antibiotics; (3) Spontaneous resolution without antibiotic therapy. The quality control strains used were Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853, provided by the Clinical Laboratory Center of the National Health and Family Planning Commission. The study was approved by the Hospital Ethics Committee (NO. 2025-L104) (In Q1 2021 and Q3 2023, our hospital did not perform statistical analysis due to insufficient S. marcescens detections).

    Strain Identification and Antimicrobial Susceptibility Testing

    Strain identification and susceptibility testing were performed using the VITEK-2 system (BioMerieux, France) and validated by the Kirby-Bauer disk diffusion method. Quality control included: (1) daily calibration with control strains (Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853); (2) retesting ≥5% of strains per batch (minimum 5 strains); (3) recalibration and data review if results deviated from CLSI 2024 standards.11 Results were interpreted as susceptible, intermediate, or resistant, with resistance rate calculated as the percentage of resistant strains among total strains. Testing was conducted by our hospital’s microbiology lab, with data managed by the Infection Control Department.

    Antibacterial Drug Usage Data

    The hospital HIS rational drug use system collects data on the use of antibacterial drugs. The WHO-defined daily dose (DDD) serves as the standard reference for calculating antimicrobial agents’ consumption in DDD units. DDDs quantify prescribing frequency, with higher values indicating stronger clinical preference for an agent. The calculation formula is as follows:

    DDDs = Annual drug total consumption/Drug DDD value.

    Data Analysis

    Data analysis used SPSS 27.0 and R 4.5.1: (1) Pearson correlation analysis (significance level α=0.05) was conducted to assess the relationship between the resistance rate of S. marcescens and DDDs of antimicrobials. (2) Multiple linear regression was used to explore independent, linear correlations. The model used S. marcescens resistance rate as the dependent variable and statistically significant antimicrobial DDDs from the previous quarter as independent variables. The stepwise method was applied for variable selection. It requires linear relationships, independent and normally distributed residuals, homoscedasticity, and no multicollinearity among independent variables. (3) The final regression model is subjected to threefold validation. Bayesian analysis used Markov Chain Monte Carlo (MCMC) simulation (10,000 iterations; weakly informative priors) in SPSS 27.0. Significance required: 95% credible interval excluding zero and posterior probability >0.95. Bootstrap Validation: Conduct 3000 BCa resamples; require original coefficient within 95% CI and CI not crossing zero. Winsorized Regression: Apply 1% bilateral winsorization, re-model. Accept if |Δβ%|<10% and p<0.05 remains significant (Δβ% = [(β_winsorized − β_original)/β_original] × 100%) and the significance level (p<0.05).

    Results

    Isolation Rate of S. marcescens

    In the past four years, 16190 strains of pathogens were detected in our hospital, among which 522 strains of S. marcescens were isolated, accounting for 3.22% of the total number of bacteria. The total χ² value was 16.365, less than the critical value (21.026), P>0.05, and there was no significant difference in the isolation rate of S. marcescens. The total number of pathogens and S. marcescens detected in each quarter are shown in Table 1. It can be seen that the detection rate of S. marcescens is 1.96%–4.35%. Of the 522 isolates of S. marcescens, 452 (86.59%) came from respiratory tract, 31 (5.94%) from blood flow, 29 (5.56%) from urine, and the remaining 10 (1.91%) from other sources. According to the statistics of the departments submitted for examination, the top 3 departments with the number of isolates of S. marcescens were 265 (50.77%) in surgery, 139 (26.63%) in ICU and 85 (16.28%) in internal medicine.

    Table 1 Detection of Serratia marcescens

    Usage of Antibacterial Drugs

    The DDDs of commonly used antimicrobial agents in the past four years are shown in Table 2.

    Table 2 The DDDs of Commonly Used Antimicrobial Agents

    β-lactam antibiotics continue to dominate, with high DDDs for piperacillin tazobactam, cefoperazone-sulbactam, and levofloxacin. Over the four years, the usage of cefuroxime, piperacillin-tazobactam, and gentamicin showed a significant upward trend.

    Drug Resistance of S. marcescens

    The isolates of S. marcescens were highly susceptible to amikacin and tigecycline, with resistance rates below 5%. They also showed susceptibility to piperacillin-tazobactam, ceftazidime, ertapenem, imipenem, meropenem, trimethoprim-sulfamethoxazole, and tigecycline, with resistance rates below 10%. Resistance to cefuroxime exceeded 90%, while resistance to amoxicillin-clavulanate remained above 60%. From 2021 to 2024, resistance rates to ceftriaxone, ceftazidime, cefoxitin, imipenem, meropenem, ertapenem, and ceftazidime decreased overall, while resistance to levofloxacin increased. For more details, as shown in Table 3.

    Table 3 Drug Resistance of Serratia marcescens

    Pearson Correlation Analysis

    Pearson correlation analysis showed that S. marcescens resistance to ceftazidime, cefoperazone-sulbactam, and imipenem correlated with the DDDs of multiple antibiotics. In contrast, resistance to amoxicillin-clavulanic acid, cefuroxime, cefoxitin, cefepime, amikacin, and levofloxacin each correlated with only one antibiotic’s DDD (Table 4).

    Table 4 Correlation of Serratia marcescens Resistance Rates (Columns) with Antibiotic DDDs (Rows)

    Table 5 Linear Regression Analysis of Serratia marcescens Resistance and DDDs

    Correlation Analysis

    To assess independent linear correlations, we utilized multiple linear regression models. Models 1–3 included multiple independent variables, while Models 4–9 incorporated single independent variables. Table 5 shows the key parameters and test values for each regression model: regression coefficients (β) quantify the impact of independent variables on the dependent variable (p<0.05 indicates significance), the F-value assesses overall model significance, VIF<5 demonstrates low multicollinearity, R² represents explanatory power, and a Durbin-Watson value near 2 confirms residual independence. Diagnostic plots for Models 1–3 are shown in Figures 1–3. The standardized residual histograms (Figures 1a, 2a and 3a) approximate a normal distribution across all groups. The normal probability (P-P) plots (Figures 1b, 2b and 3b) demonstrate points clustering closely along the diagonal line, confirming the normality of residuals. The residual scatter plots (Figures 1c, 2c and 3c) show evenly distributed points within the ±3 range for all groups, indicating homoscedasticity and absence of influential outliers.

    Figure 1 Regression diagnostic plots for Model 1. (a). Histogram of standardised residuals (normality assessment), (b). Normal Q-Q plot (residual normality verification), (c). Scatterplot of Residuals versus fitted values (homoscedasticity assessment).

    Figure 2 Regression diagnostic plots for Model 2. (a). Histogram of standardised residuals (normality assessment), (b). Normal Q-Q plot (residual normality verification), (c). Scatterplot of Residuals versus fitted values (homoscedasticity assessment).

    Figure 3 Regression diagnostic plots for Model 3. (a). Histogram of standardised residuals (normality assessment), (b). Normal Q-Q plot (residual normality verification), (c). Scatterplot of Residuals versus fitted values (homoscedasticity assessment).

    Robustness Verification

    The final regression model is subjected to threefold validation (Table 6). Multiple robustness validations consistently demonstrated negative correlations for IPM-RR/GEN, CFS-RR/GEN, CFX-RR/TZP, and FEP-RR/CXM. Bayesian posterior means were less than zero (95% HPD excluding zero), Bootstrap confidence intervals were entirely negative, and Winsorized regression coefficients were highly consistent with the original model, with minimal changes such as +0.41% for IPM-RR/GEN.

    Table 6 Robustness Verification

    The robustness validation for the AMC-RR/MEM and LVX-RR/GEN groups supports the original positive trend. Bayesian analysis shows posterior means significantly greater than zero, and Winsorized regression reinforces this with positive coefficients and confidence intervals excluding zero. Bootstrap results are borderline, primarily reflecting insufficient estimation precision rather than a change in effect direction. All three validation methods have consistent positive point estimates, with effect sizes exceeding the clinical threshold (|β| > 0.02), confirming the biological validity of the positive trend.

    The triple validation showed consistent negative trends for CAZ-RR/TZP, CXM-RR/TZP, and AMK-RR/TZP. Bayesian and Winsorized regression results were statistically significant, while Bootstrap analysis results were borderline, reflecting limited estimation accuracy due to the sample size. The effect sizes were small (|β| < 0.005) and did not reach the clinical threshold (|β| > 0.02), but the consistent trends warrant validation through enhanced analytical approaches in future investigations.

    Discussion

    S. marcescens, first identified in 1819 by Venetian pharmacist Bartolomeo Bizio, was initially seen as a low-pathogenicity environmental saprophyte. However, its pathogenic potential was revealed in the 1950s through fatal urinary tract infections. It can cause infections in multiple systems, including respiratory, neurological, abdominal, cardiovascular, musculoskeletal, and ocular.12 Recently, S. marcescens has gained attention as an emerging pathogen worldwide, provoking infections and outbreaks in debilitated individuals, particularly newborns and patients in ICU.13 S. marcescens resists disinfectants, persists in the environment, and colonizes medical devices and surfaces. Its invasiveness and antimicrobial resistance make it a major hospital-acquired pathogen.

    In the past four years, a total of 522 strains of S. marcescens were detected in our hospital, accounting for 3.22% of the total bacterial count in the hospital. The clinical isolated samples of S. marcescens mainly come from the respiratory tract, accounting for 86.59%, and the proportion of samples from blood, urine, secretions, ducts, etc. is relatively small. This suggests that when the immune system is weakened, the respiratory tract is a common site for S. marcescens infection, calling for clinical prevention and control. This finding aligns with relevant literature.14,15 The presence of S. marcescens can exacerbate the condition of patients with compromised immune systems.16 Furthermore, studies have indicated that this bacterium may lead to severe acute infections, resulting in respiratory distress even in individuals with normal immune function.17 S. marcescens is mainly found in surgical and intensive care unit (ICU), emphasizing the need for better infection control. This aligns with Chen Jian et al’s findings.18 Major risk factors include serious underlying diseases, major surgeries, ventilator use, tracheostomies, invasive exams, and prolonged high-dose antibiotics, which are common among ICU and surgical patients.

    S. Marcescens contains R factors or R plasmids that can carry multiple resistance genes, leading to its natural resistance to first generation cephalosporins, ampicillin, and macrolides.18 The study showed that S. marcescens had a resistance rate of over 90% to cefuroxime and a high resistance rate to amoxicillin-clavulanic, consistent with Li Jin et al.19 The isolated strains of S. marcescens showed sensitivity to cefoperazone-sulbactam, cefepime, ertapenem, imipenem, meropenem, amikacin, trimethoprim-sulfamethoxazole, and tigecycline, with a resistance rate of less than 10%, consistent with previous reports.20 S. marcescens shows high sensitivity to amikacin and tigecycline, with resistance rates below 5%. Compared with the national data of bacterial resistance monitoring in CHINET in 2023,4 the resistance rate of 522 strains of S. marcescens to cefuroxime is slightly higher than the national average level. The drug sensitivity results to ceftriaxone, ceftazidime, levofloxacin, amikacin, trimethoprim-sulfamethoxazole, and tigecycline are relatively consistent with the national data, while the resistance rate to cefepime, cefoxitin, cefoperazone-sulbactam, ertapenem, imipenem and meropenem is lower than the national level. Clinically, third-generation cephalosporins, aminoglycosides, and quinolones treat S. marcescens infections, with carbapenems for severe MDR strains. Overuse of carbapenems has coincided with an increasing detection rate of carbapenem-resistant S. marcescens, hindering control and treatment.21 Notably, studies demonstrate that S. marcescens exhibits intrinsic resistance to polymyxins. Studies show S. marcescens is naturally resistant to polymyxins,6 if it also produces carbapenemases, treatment gets harder.22 Clinicians should choose antibiotics wisely to control and prevent outbreaks of S. marcescens infections.

    Reports indicate that developing a new antimicrobial drug from research to clinical use generally takes 5 to 10 years. In contrast, bacteria can develop resistance in as few as 2 years.23 Thus, slowing bacterial resistance is just as critical as enhancing antimicrobial drug efficacy. The study showed resistance to cefoperazone-sulbactam and imipenem had independent negative linear relationships with gentamicin DDDs; resistance to cefoxitin correlated negatively with piperacillin-tazobactam DDDs; resistance to cefepime showed a negative association with cefuroxime DDDs. These four relationships were strongly supported by consistent results from Bayesian, Bootstrap, and Winsorized regression. In addition, amoxicillin-clavulanic resistance positively correlated with meropenem DDDs, and levofloxacin resistance positively correlated with gentamicin DDDs. These positive trends were supported by triple robustness testing. Concurrently, analyses of S. marcescens isolates revealed negative linear correlations of resistance to ceftazidime, cefuroxime, and amikacin with piperacillin-tazobactam DDDs. Although these associations were directionally consistent across all robustness validation methods, their effect sizes fell below prespecified clinical thresholds, warranting further studies to explore these potential relationships. The recurrent involvement of piperacillin-tazobactam in resistance patterns warrants attention. Existing studies have shown that S. marcescens has the ability to form biofilms, which can hinder antibiotic penetration, alter bacterial metabolic states, and thus enhance bacterial drug resistance.24 Although research on S. marcescens is limited, studies on other Gram-negative bacteria can offer insights. Research has demonstrated that use of piperacillin-tazobactam reduces rates of ceftazidime-resistant Klebsiella pneumoniae, which is similarly a member of the Enterobacteriaceae family.25 The mechanism may be related to biofilm inhibition. Studies have shown that sub-minimal inhibitory concentration (sub-MIC) concentrations of piperacillin-tazobactam can effectively reduce biofilm formation by inhibiting the adhesiveness and motility of Escherichia coli,26 and reduce Pseudomonas aeruginosa biofilm formation in a dose-dependent manner, with the strongest inhibitory effect at MIC/2 concentration.27 Whether increased use of piperacillin-tazobactam reduces S. marcescens resistance to ceftazidime by inhibiting biofilm-mediated resistance barriers remains unclear, and relevant mechanisms require further research through constructing in vitro biofilm models and conducting in vivo antibacterial experiments. Zhao Ningqiu et al28 found that different antimicrobial agents may share resistance mechanisms, such as overexpression of efflux pumps, changes in cell membrane permeability, inactivation of antimicrobial compounds, and modification of their targets. These mechanisms can cause cross-resistance or co-resistance between drugs.

    From a biological perspective, these associations may involve complex mechanisms. Negative correlations may reveal potentialistic synerg inhibitory effects between antimicrobial drugs, suggesting that increased use of certain drugs can suppress the rise in resistance to target drugs, indicating that rational combination or rotation of antimicrobial drugs can effectively reduce resistance risks. Positive correlations suggest the presence of cross-resistance or co-resistance mechanisms, implying that increased use of one class of antimicrobial drugs may simultaneously or subsequently lead to increased resistance to other structurally or functionally similar drugs. Hospitals should optimize antibiotic management strategies, including targeted antibiotic rotation, limiting high-risk drug use, and adhering to data-driven guidelines. These measures can improve patient care, curb the spread of S. marcescens resistance, and preserve the long-term efficacy of essential antimicrobials.

    The resistance mechanisms of S. marcescens are complex, as it inherently exhibits resistance to multiple antibiotics, such as polymyxin and cephalothin. It can produce specific β-lactamases induced by drugs, include cephalosporinase (AmpC), extended-spectrum β-lactamases (ESBLs), carbapenemases, etc. These enzymes are capable of hydrolyzing the β-lactam ring of β-lactam antibiotics, rendering the antibiotics inactive and thus enabling the bacteria to develop resistance to the corresponding antibiotics. Research on carbapenem resistance mainly focuses on serine carbapenemases, such as KPC enzyme, OXA-48, and metallo-enzymes.29 The resistance genes encoding carbapenemases in S. marcescens are mainly KPC-2,30 NDM-1,31 IMP-1, IMP-10, and it may carry multiple resistance genes, presenting a multi-drug-resistant phenotype.32 Relevant reports have shown that the synergistic effect of high-yield chromosome-mediated loss of AmpC enzyme and porin resulting in decreased permeability can also cause drug resistance to carbapenilic antibiotics.33 Moreover, reduced outer membrane permeability and overactive efflux pumps in the inner membrane are key contributors to increased drug resistance in S. marcescens.34

    Currently, there is a scarcity of global data regarding how S. marcescens develops resistance in response to antimicrobial exposure. This aspect remains a relatively unexplored area within S. marcescens research. Our study used multiple linear regression to analyze the independent, linear correlation between S. marcescens resistance rates and antibiotic use one quarter in advance. We also validated our results using three methods: Bayesian, which enhances reliability; Bootstrap, which improves model adaptability; Winsorized regression, which controls the impact of outliers. Our quarterly time-series design (longitudinal study) is more effective at capturing macro trends compared to cross-sectional studies.9,35 However, the limited observation period restricts the number of data points in similar longitudinal studies.36,37 To compensate, we optimized our methodology through triple validation to ensure model stability. Future research could further investigate the impact of continuous antibiotic use on resistance through randomized controlled trials and explore resistance mechanisms. We hope this study offers a unique perspective to guide antibiotic management.

    The rapid emergence and spread of resistant bacteria in hospitals are due to multiple factors, including improper antimicrobial use creating high selective pressure, inconsistent infection control measures leading to patient-to-patient transmission, inter-hospital transfer of resistance, and complex links between resistance and antimicrobial use.38 Clinicians should strictly control antimicrobial use, standardize dosages and durations, and avoid misuse that can trigger resistance. Tailoring treatments based on infection type and susceptibility results can reduce resistance risks. Healthcare facilities need to enhance resistance monitoring, strengthen multidisciplinary collaboration, and enforce strict antimicrobial use regulations. Regular analysis of resistance data can guide rational prescribing. Infection control measures, including “one use, one disinfection”, isolating patients with resistant infections, and using dedicated equipment, are essential. Hand hygiene training and rapid response to clustered cases are crucial to break transmission chains. Understanding the correlation between S. marcescens resistance rates and antimicrobial use is crucial for clinical resistance control.

    Conclusion

    Over the past four years, our hospital identified 522 S. marcescens strains (3.22% of all isolates), with 86.59% from respiratory samples. This indicates that the respiratory tract is a common site for S. marcescens infection, necessitating clinical prevention and control measures. Most strains were sensitive to cefoperazone-sulbactam, cefepime, ertapenem, imipenem, meropenem, amikacin, trimethoprim-sulfamethoxazole, and tigecycline, with resistance rates <10%. Our study found significant correlations between S. marcescens resistance and antibiotic usage. Negative correlations showed that strategic antimicrobial use can reduce resistance rates, while positive correlations indicated higher co-resistance risks. These results highlight the need for hospitals to implement antibiotic rotation, limit high-risk drug use, and enhance supervision to improve patient outcomes and curb resistance.

    Highlights

    1. This study is the first to deeply investigate the macro-level quantitative relationship between S. marcescens drug resistance and antibiotic use, filling a research gap in this area.

    2. Using multiple linear regression models, we identified independent linear correlations between S. marcescens resistance rates and the use of various antibiotics.

    3. Our findings provide a scientific basis for rational and optimized antibiotic use in clinical practice, helping to delay the emergence of resistant bacteria and combat the threat of antibiotic resistance.

    Ethics Approval

    This study was reviewed and approved by the Ethics Committee of Wenzhou Hospital of Integrated Traditional Chinese and Western Medicine (NO. 2025-L104). Informed consent was obtained from the patient.

    Acknowledgments

    We are grateful to Dr Ren-ai Xu for his invaluable guidance and support throughout this research. We also thank Ms Leping Liu for her helpful comments and suggestions.

    Disclosure

    The authors report no conflicts of interest in this work.

    References

    1. Duan XK, Liu WF, Xiao YY, et al. Exploration of the feasibility of clinical application of phage treatment for multidrug-resistant Serratia marcescens-induced pulmonary infection. Emerg Microbes Infect. 2025;14(1):2451048. doi:10.1080/22221751.2025.2451048

    2. Hanczvikkel A, Tóth Á, Kopcsóné Németh IA, et al. Nosocomial outbreak caused by disinfectant-resistant Serratia marcescens in an adult intensive care unit, Hungary, February to March 2022. Euro Surveill. 2024;29(26):2300492. doi:10.2807/1560-7917.ES.2024.29.26.2300492

    3. Muyldermans A, Crombé F, Bosmans P, et al. Serratia marcescens outbreak in a neonatal intensive care unit and the potential of whole-genome sequencing. J Hosp Infect. 2021;111:148–154. doi:10.1016/j.jhin.2021.02.006

    4. Guo Y, Hu FP, Zhu DM, et al. CHINET surveillance of bacterial resistance in China: 2023 report. Chin J Infect Chemother. 2024;24(6):627–637.

    5. Hu FP, Guo Y, Zhu DM, et al. CHINET surveillance of bacterial resistance in China: 2019 report. Chin J Infect Chemother. 2020;20(3):233–243.

    6. Ferreira RL, Rezende GS, Damas MSF, et al. Characterization of KPC-producing Serratia marcescens in an intensive care unit of a Brazilian tertiary hospital. Front Microbiol. 2020;11:956. doi:10.3389/fmicb.2020.00956

    7. Bolourchi N, Noori Goodarzi N, Giske CG, et al. Comprehensive pan-genomic, resistome and virulome analysis of clinical OXA-48 producing carbapenem-resistant Serratia marcescens strains. Gene. 2022;822:146355. doi:10.1016/j.gene.2022.146355

    8. Yan Y, Sun WG. Impact of antimicrobial consumption frequency on Staphylococcus aureus resistance rates in the following quarter. Chin J Pharmacoepidemiol. 2018;27(6):382.

    9. Bronzwaer SL, Cars O, Buchholz U, et al. A European study on the relationship between antimicrobial use and antimicrobial resistance. Emerg Infect Dis. 2002;8(3):278–282. doi:10.3201/eid0803.010192

    10. Shang H, Wang YS, Shen ZY. National Guide to Clinical Laboratory Procedures. 4th ed. Beijing: People’s Medical Publishing House; 2015:560–811.

    11. Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing: thirty-fourth informational supplement. In: Document. Wayne, PA: CLSI; 2024:M100–S34.

    12. Iovene MR, Pota V, Galdiero M, et al. First Italian outbreak of VIM-producing Serratia marcescens in an adult polyvalent intensive care unit, August-October 2018: a case report and literature review. J Hosp Infect. 2019;101(4):431–437. doi:10.12998/wjcc.v7.i21.3535

    13. Tavares-Carreon F, De Anda-Mora K, Rojas-Barrera IC, et al. Serratia marcescens antibiotic resistance mechanisms of an opportunistic pathogen: a literature review. PeerJ. 2023;11:e14399. doi:10.7717/peerj.14399

    14. Zhang Y, Zhou WQ, Li J, et al. Resistance phenotype analysis of 53 carbapenem-resistant Serratia marcescens isolates. J Clin Transfus Lab Med. 2021;23(1):87–90.

    15. Simsek MD. Determination of the antibiotic resistance rates of Serratia marcescens isolates obtained from various clinical specimens. Niger J Clin Pract. 2019;22(1):125–130. doi:10.4103/njcp.njcp_362_18

    16. Muyldermans A, Crombé F, Bosmans P, et al. Serratia marcescens outbreak in a neonatal intensive care unit and the potential of whole-genome sequencing. J Hosp Infect. 2020;104(4):438–445. doi:10.1016/j.jhin.2021.02.006

    17. Ruiz-Sada P, Escalante M, Lizarralde E. Severe acute infection due to Serratia marcescens causing respiratory distress in an immunocompetent adult. Rom J Intern Med. 2016;54(2):134–136. doi:10.1515/rjim-2016-0013

    18. Chen J, Xu CQ, Tu HX. Drug resistance analysis of Serratia marcescens in hospital infections. Chin J Health Lab Technol. 2009;19(12):2889–2990.

    19. Li J, Zhang XY. Epidemiological characteristics and antimicrobial resistance of Serratia marcescens in a tertiary hospital from 2017 to 2023. Chin J Infect Control. 2025;24(1):52–57.

    20. Zhang ZX, Wu WH. Distribution and drug resistance analysis of 56 Serratia marcescens isolates. Chin J Disinfect. 2013;30(12):1134–1137.

    21. Yuan Y, Chen JZ, Pan YP, et al. Carbapenem-resistant Serratia marcescens: resistance genes and molecular epidemiology. Acta Univ Med Anhui. 2019;54(1):33–38.

    22. Ma JL, Liu ZZ, Li DJ, et al. Molecular epidemiology of carbapenemase-producing Serratia marcescens infections in a neurosurgical ICU. Chin J Nosocomiol. 2020;30(3):393–397.

    23. Zou J, Li N. Bacterial resistance mechanisms and control strategies. Basic Med Forum. 2006;10(1):1052–1053.

    24. Erdal B, Yalınay M, Elmas Ç, et al. Investigation of pseudomonas aeruginosa biofilm formation and quorum sensing genes in piperacillin/tazobactam and ciprofloxacin sub-minimal inhibitory concentrations. Mikrobiyol Bul. 2020;54(4):547–558. doi:10.5578/mb.70087

    25. Hsueh PR, Chen WH, Luh KT. Relationships between antimicrobial use and antimicrobial resistance in gram-negative bacteria causing nosocomial infections from 1991–2003 at a university hospital in Taiwan. Int J Antimicrob Agents. 2005;26(6):463–472. doi:10.1016/j.ijantimicag.2005.08.016

    26. Song J, Hong H, Feng W, et al. Effect of subinhibitory concentrations of piperacillin/tazobactam on biofilm formation ability of Escherichia coli. Chongqing Med. 2018;47(3):368–370.

    27. Ray C, Shenoy AT, Orihuela CJ, et al. Killing of Serratia marcescens biofilms with chloramphenicol. Ann Clin Microbiol Antimicrob. 2017;16(1):19. doi:10.1186/s12941-017-0192-2

    28. Zhao NQ, Liu C, Xue YX, et al. Cross-resistance and co-resistance of bacteria to antimicrobial compounds. Chin J Antibiot. 2021;46(1):11–19.

    29. Wu HO, Zeng WL, Dai YY, et al. Molecular mechanisms of carbapenemase-producing Serratia marcescens. Chin J Health Lab Technol. 2022;32(2):170–173.

    30. Rodriguez C, Brengi S, Caceres MA, et al. Polyclonal dissemination of KPC-2 in Serratia marcescens, including a clone with epidemic behavior in the nosocomial niche. Int J Antimicrob Agents. 2017;49(5):657–658. doi:10.1016/j.ijantimicag.2017.03.003

    31. Phan HTT, Stoesser N, Maciuca IE, et al. Illumina short-read and MinION long-read WGS to characterize the molecular epidemiology of an NDM-1 Serratia marcescens outbreak in Romania. J Antimicrob Chemother. 2018;73(3):672–679. doi:10.1093/jac/dkx456

    32. Wang JM, Zhang HQ, Zhang YH, et al. Resistance mechanisms and homology analysis of carbapenem-resistant Serratia marcescens in 12 burn patients. Chin J Health Lab Technol. 2017;27(9):1354–1356,1359.

    33. Guo P, Qiao Y, Zhang HT, et al. Drug resistance and imipenem resistance mechanisms of clinical Serratia marcescens isolates. Chin J Infect Chemother. 2017;17(2):187–191.

    34. Montagnani C, Cocchi P, Lega L, et al. Serratia marcescens outbreak in a neonatal intensive care unit: crucial role of implementing hand hygiene among external consultants. BMC Infect Dis. 2015;15:11. doi:10.1186/s12879-014-0734-6

    35. Olesen SW, Barnett ML, MacFadden DR, et al. The distribution of antibiotic use and its association with antibiotic resistance. Elife. 2018;18(7):e39435. doi:10.7554/eLife.39435

    36. Bruyndonckx R, Hens N, Aerts M, et al. Exploring the association between resistance and outpatient antibiotic use expressed as DDDs or packages. J Antimicrob Chemother. 2015;70(4):1241–1244. doi:10.1093/jac/dku525

    37. Meyer E, Gastmeier P, Deja M, et al. Antibiotic consumption and resistance: data from Europe and Germany. Int J Med Microbiol. 2013;303(6–7):388–395. doi:10.1016/j.ijmm.2013.04.004

    38. White RL, Friedrich LV, Mihm LB, et al. Assessment of the relationship between antimicrobial usage and susceptibility: differences between the hospital and specific patient-care areas. Clin Infect Dis. 2000;31(1):16–23. doi:10.1086/313916

    Continue Reading

  • Devil Wears Prada 2 to Spider-Man 4: is the Hollywood hype machine starting too early? | Film

    Devil Wears Prada 2 to Spider-Man 4: is the Hollywood hype machine starting too early? | Film

    Do you want to know who might be starring, cameoing or in some way appearing in the upcoming sequel to The Devil Wears Prada? Well, you might not have a choice. Ever since filming on the Anne Hathaway/Meryl Streep reunion commenced in New York a few weeks ago, set photos and accompanying news items have become plentiful (don’t click that link if you don’t want to know the rumors about a major star). Of course, gossip pages breathlessly reporting casting rumors are not sanctioned by a movie studio, like a teaser, poster, trailer or announcement video. At the same time, the movie hype machine has been blurring the line between genuine marketing and tweet-grade micro-teases all summer long.

    For example: is there a teaser trailer for next summer’s Spider-Man: Brand New Day? It’s the next Marvel movie after this July’s Fantastic Four reboot, and plenty of summer blockbusters-to-be have called their shots with a teaser a year in advance, so it would make sense to get a little something out there. A teaser is kinda-sorta what Sony posted a few weeks ago, only it was hyped up as a “suit reveal”, which seems like an attempt to redress a 20-second teaser in literal new clothes. It also seems like a backfire if the strategy involves anyone beyond the most diehard 5% of movie-watchers realizing or caring that Tom Holland walks up to the camera wearing a Spider-Man costume that is slightly different than the six other Spider-Man costumes he’s donned in other appearances. There is also a behind-the-scenes video of a movie that’s barely started filming. It recalls the absurdly lengthy “chair reveal” of cast members from Avengers: Doomsday parceled out over five hours, which is getting its own sequel soon. (Not everyone was announced, of course. Not least because the movie’s screenplay was not finished at the time of that announcement and filming commencement.)

    Even Christopher Nolan, who has generally mastered the art of keeping his highly anticipated movies to the realm of teasers and trailers – a teaser for his Imax-shot version of The Odyssey emerged just about a year before its planned July 2026 release – succumbed to the hype machine when that Odyssey teaser was accompanied by tickets for select showtimes going on sale. That’s right, if you want to plan a movie outing 50 weeks from now, you can see about booking an Odyssey ticket at your local Imax (if applicable). Or rather, you could have if your AMC app didn’t crash; most of those showings sold out quickly.

    To some extent, this is business as usual. There’s been year-out hype for movies for literal decades at this point; Alien 3 had a teaser trailer that infamously teased a movie with a completely different premise (Aliens on Earth?!). Some of it is also just the studios attempting to compete with the speed and ferocity of digital-era hype, to be heard above the constant buzz of phone notifications over virtually nothing (as well as the semi-regular global calamity). After all, the days of getting by on trailers, posters and TV ads alone are long gone, right?

    Maybe not. The biggest global opening of the past weekend was the horror movie Weapons, which exceeded expectations for its launch based largely off of … trailers that explained the premise of the movie. Obviously there’s more to a marketing campaign than that. In the spring, the film launched a website mocked up to look like news postings about Maybrook, the fictional town in the movie, with shades of The Blair Witch Project from 1999. But the baseline of the movie’s marketing was a couple of trailers that showed a lot of creepy imagery but said little about the multi-character plot beyond what the movie reveals in its first two minutes: one night, a group of third-graders from the same class mysteriously rise from their beds, run outside, and disappear into the night. The movie is about figuring out what happened to them; no casting announcements, announcement videos, or greetings from the set needed. For well over a year, about the only thing anyone repeated about this movie in the Hollywood trade publications was that it was a “horror epic” with some elements in common with another 1999 classic, Paul Thomas Anderson’s Magnolia.

    Of course, Weapons operated with the luxury of less attention from both its studio and the outside press. It’s not the kind of superhero movie or high-profile sequel that gets clicks a year before it comes out. But this strategy isn’t only employed for smaller films. Disney’s Lilo & Stitch remake, the second-biggest global hit of the year so far, certainly had plenty of promotion, tie-ins, brand activations and so on. But it successfully waited until much closer to its release date, putting out a couple of traditional low-footage teasers at the end of 2024, and a full trailer a few months before the movie’s release in 2025. Sinners, an ambitious movie with an audience more in line with Weapons, also managed to avoid selling much-hyped nothingburgers, and was able to take audiences by surprise.

    Next year’s crop of much-hyped sequels will still probably draw crowds. Streep, Hathaway, Nolan, Spider-Man and whoever is in the Avengers these days all have plenty of fans. But the rush to blow up what would normally be lines in a press release to an epic and ceremonious scale still feels sweaty – maybe sweatier, even, as if companies including Disney are desperate to project the image of day-stopping news when another name of an old X-Men cast member turns up on a chair. As recently as the release of Deadpool & Wolverine, the company prized spoiler protection, with almost none of the movie’s various in-joke cameos signaled ahead of time. That seemed like good marketing, but maybe it was just fear of the fact that many of those cameos were intentionally expectations-undermining jokes, rather than mega-applause moments. It’s harder to tell in retrospect because the hype machine has spent so much time accidentally eliminating the difference between teasers, spoilers and piles of random crumbs.

    Continue Reading

  • Accedo to Showcase Identity Management for Accedo One at IBC2025

    Accedo to Showcase Identity Management for Accedo One at IBC2025

    STOCKHOLM, Sweden—Accedo, a global provider of video streaming software and services, is launching identity management functionality for its Accedo One SaaS platform and will showcase it during IBC2025, Sept. 12-15, at the RAI Amsterdam Convention Center.

    Accedo One supports user registration and authentication, something previously only possible through third-party integrations. This function makes it easier to manage registration and authentication on a single platform.

    “Identity management is a critical building block for any video service. By bringing it natively into Accedo One, we are making it easier than ever for our customers to manage their subscribers, helping them to build, scale and evolve their video services,” Accedo CEO Markus Hejdenberg said.

    The upcoming subscription management suite will enable media companies to manage the full subscriber life cycle within Accedo One—from sign-up to billing, retention and churn management. Accedo One will continue to support integrations with external partners in the identity and subscription space, giving customers the option to choose the solution that best fits their strategy, the company said.

    See Accedo at IBC2025 in Stand 5.F80.

    Continue Reading

  • Researchers Use Next-Level Pixel-Particle Analogy to Remove Noise from Medical Images

    Researchers Use Next-Level Pixel-Particle Analogy to Remove Noise from Medical Images

    Aug. 12, 2025 – Medical imaging methods such as ultrasound and MRI are often affected by background noise, which can introduce blurring and obscure fine anatomical details in the images. For clinicians who depend on medical images, background noise is a fundamental problem in making accurate diagnoses.

    Methods for denoising have been developed with some success, but they struggle with the complexity of noise patterns in medical images and require manual tuning of parameters, adding complexity to the denoising process.

    To solve the denoising problem, some researchers have drawn inspiration from quantum mechanics, which describes how matter and energy behave at the atomic scale. Their studies draw an analogy between how particles vibrate and how pixel intensity spreads out in images and causes noise. Until now, none of these attempts directly applied the full-scale mathematics of quantum mechanics to image denoising.

    In a paper in the American Institute of Physics’ AIP Advances, by AIP Publishing, researchers from Massachusetts General Hospital, Harvard Medical School, Weill Cornell Medicine, GE HealthCare and Université de Toulouse took a particle-pixel analogy to the next level.

    “While quantum localization is a well-established phenomenon in physical materials, our key innovation was conceptualizing it for noisy images — translating the physics literally, not just metaphorically,” author Amirreza Hashemi said. “This foundational analogy didn’t exist before. We’re the first to formalize it.”

    A central concept in the math describing matter and energy, localization is used to explain how particles vibrate in a space. Vibrations that stay confined are considered localized, while vibrations that spread out are diffused. Similarly, pixel intensity, or brightness, in a clear image can be considered localized, while noisy patterns in an image can be considered diffused.

    The authors apply the same mathematics that describes the localization of particle vibrations in the surrounding physical space to decipher the localization of pixel intensity in images. In this way, they can separate the noise-free “signal” of the anatomical structures in the image from the visual noise of stray pixels.

    “The main aspect was developing an algorithm that automatically separates the localized (signal) and nonlocalized (noise) components of pixels in an image by exploiting their distinct behaviors,” Hashemi said.

    The researchers’ direct application of the physics and mathematics of particles also eliminated the need to manually fine-tune parameters in denoising algorithms, which Hashemi said is a major hindrance in traditional approaches.

    “Our method leverages physics-driven principles, like localization and diffusive dynamics, which inherently separate noise from signal without expensive optimization,” Hashemi said. “The algorithm just works by design, avoiding brute-force computations.”

    Their method has applications not only in medical image denoising, but across quantum computing, too.

    “Our physics-driven framework aligns with the computational primitives of quantum systems, offering a potential performance advantage as quantum computing scales.”

     

    The article “A novel perspective on denoising using quantum localization with application to medical imaging” is authored by Amirreza Hashemi, Sayantan Dutta, Bertrand Georgeot, Denis Kouamé and Hamid Sabet. It can be accessed at https://doi.org/10.1063/5.0267924.


    Continue Reading

  • Claude Sonnet 4 now supports 1M tokens of context \ Anthropic

    Claude Sonnet 4 now supports 1M tokens of context \ Anthropic

    Claude Sonnet 4 now supports up to 1 million tokens of context on the Anthropic API—a 5x increase that lets you process entire codebases with over 75,000 lines of code or dozens of research papers in a single request.

    Long context support for Sonnet 4 is now in public beta on the Anthropic API and in Amazon Bedrock, with Google Cloud’s Vertex AI coming soon.

    Longer context, more use cases

    With longer context, developers can run more comprehensive and data-intensive use cases with Claude, including:

    • Large-scale code analysis: Load entire codebases including source files, tests, and documentation. Claude can understand project architecture, identify cross-file dependencies, and suggest improvements that account for the complete system design.
    • Document synthesis: Process extensive document sets like legal contracts, research papers, or technical specifications. Analyze relationships across hundreds of documents while maintaining full context.
    • Context-aware agents: Build agents that maintain context across hundreds of tool calls and multi-step workflows. Include complete API documentation, tool definitions, and interaction histories without losing coherence.

    API pricing

    To account for increased computational requirements, pricing adjusts for prompts over 200K tokens:

    Input Output
    Prompts ≤ 200K $3 / MTok $15 / MTok
    Prompts > 200K $6 / MTok $22.50 / MTok
    Claude Sonnet 4 pricing on the Anthropic API

    When combined with prompt caching, users can reduce latency and costs for Claude Sonnet 4 with long context. The 1M context window can also be used with batch processing for an additional 50% cost savings.

    Customer spotlight: Bolt.new

    Bolt.new transforms web development by integrating Claude into their browser-based development platform.

    “Claude Sonnet 4 remains our go-to model for code generation workflows, consistently outperforming other leading models in production. With the 1M context window, developers can now work on significantly larger projects while maintaining the high accuracy we need for real-world coding,” said Eric Simons, CEO and Co-founder of Bolt.new.

    Customer spotlight: iGent AI

    London-based iGent AI is advancing the field of software development with Maestro, an AI partner that transforms conversations into executable code.

    “What was once impossible is now reality: Claude Sonnet 4 with 1M token context has supercharged autonomous capabilities in Maestro, our software engineering agent at iGent AI. This leap unlocks true production-scale engineering—multi-day sessions on real-world codebases—establishing a new paradigm in agentic software engineering,” said Sean Ward, CEO and Co-founder of iGent AI.

    Get started

    Long context support for Sonnet 4 is now in public beta on the Anthropic API for customers with Tier 4 and custom rate limits, with broader availability rolling out over the coming weeks. Long context is also available in Amazon Bedrock, and is coming soon to Google Cloud’s Vertex AI. We’re also exploring how to bring long context to other Claude products.

    To learn more about Sonnet 4 and the 1M context window, see our documentation and pricing page.

    Continue Reading

  • Pakistani products that deserve a spot in your cart this August 14 – Sponsored

    Pakistani products that deserve a spot in your cart this August 14 – Sponsored

    Bazaar brings people the best products across more than 30 categories delivered to their doorstep the very next day.

    Pakistan’s Independence Day is more than just the flag and fireworks, it’s a celebration of what makes Pakistan special and what better way to celebrate Pakistan than with the best homegrown products that we use in our homes every day.

    From nostalgic cold drinks to traditional spices, your Independence Day grocery shopping is a great way to support local brands while celebrating freedom with friends and family. Whether you are hosting a dinner or planning a picnic, here are Pakistani products worth adding to your basket this Independence Day not just for patriotism but for taste, quality and convenience. The best part? You can get all three from Bazaar.

    Drinks that taste like home

    No gathering is complete without our favourite Pakistani drinks to wash down the delicious food. Instead of reaching for imported products, go for local favourites like Pakola, Cola Next or Shehzan juices (people prefer the mango flavour, but apple is popular too). These drinks are refreshing and full of nostalgia, perfect for summer BBQs and dinner parties.

    Why add to cart:

    ● Refreshing and proudly Pakistani

    ● Great value for bulk/party sizes

    ● Crowd-pleasers across generations

    Best spices for the holiday

    BBQ parties and picnics are a great way to bring friends and family together to celebrate Independence Day and local brands have all the best masalas you need to get the grill going. From Shan Masala blends like Shan Chicken Tikka Masala or Shan Chapli Kabab Masala to National Masala favourites such as National Chicken Tikka Masala, Pakistani brands know what their customers want when it comes to their BBQ. And if your household isn’t into barbeque and you’re going for other dishes, other masalas you can try include Shan Haleem Masala, Shan Nihari Masala or Shan Karahi Masala for a better variety of foods.

    Why add to cart:

    ● Made for local recipes (tikka, boti and more)

    ● Trusted blends, no guesswork needed

    ● Great for both first-timers and pros

    Snacks for evening tea

    Whether you’re hosting guests or watching fireworks, snacking is a national pastime. This Independence Day, stock up on crowd-pleasing crunchy treats including Noms Nachos or Super Crisp chips for classic flavour. For biscuits, everyone’s favourites these days (and back in the day) include Digestive Biscuits, a classic option that tastes great on its own or with a cup of chai. And if you want to go for sugar free biscuits, Bisconni Digestive Sugar Free Biscuit is a great alternative for those who want a slightly healthier option. And for the kids, if you’re looking for something sweeter like chocolate biscuits, Chocolate Sandwich Biscuits are a classic favourite across generations.

    Why add to cart:

    ● Perfect for serving guests, picnics or road trips

    ● Affordable and easy to buy in bulk

    ● Loved by both kids and grown-ups

    Ready-to-cook frozen food for quick meals

    We all want to make sure that as well as getting ready for holiday gatherings we actually enjoy our days off. This is where frozen food helps as a great way to speed up cooking and get to the fun part. Dawn frozen foods make it easy with handy options like chicken samosas, spring rolls and crispy parathas. Menu frozen food is a favourite for hearty meals with plain parathas, stuffed parathas and shami kababs. For something crispy and full of flavour Sabroso frozen food is known for family friendly choices like chicken nuggets, seekh kababs and chicken poppers which are perfect for kids plates. With all three brands offering quality, variety and quick cooking stocking up means you are ready for any August 14 gathering without extra kitchen stress.

    Why add to cart:

    ● Less prep, more time to celebrate

    ● Tasty and consistent

    ● Freezer-friendly and portioned for sharing

    Tea and milk for every type of celebration

    No matter the occasion, tea is a must. And what better way to end your holiday than with trusted local tea favourites like Tapal, Brooke Bond and Vital. Pair them with rich, locally produced milk from Nurpur or Haleeb to get that perfect cup of mixed tea.

    Why add to cart:

    ● Essential for serving guests or prepping for dinner

    ● Widely loved — every home runs on tea

    ● Great bundled with snacks, biscuits and rusks

    Cleaning supplies to get ready for gatherings

    Hosting a big celebration? You’ll want to have dependable cleaning essentials on hand and a few local brands stand out. Tez Clean offers value-packed options with their dishwashing bars that will keep your dishes sparkled at an affordable price. You can also try Bazaar Select dishwashing liquid as a more convenient option that will effectively tackle grease, which is much needed after a long day of hosting. Lastly, you can also get your hands on Sufi Safon, a classic with trusted household staples in all forms including powder, bars and a great dishwashing liquid formula.

    Why add to cart:

    ● Efficient cleanups after the guests are gone

    ● Reliable results for all your dishes

    ● Affordable products that get the job done

    Tissues for easy cleanup

    When you’re serving a crowd or eating outdoors, reliable tissues are a necessity. Premium options like Rose Petal Tissue are well-known for being reliable and soft. Other great options that you may not have heard of are Papia tissue and Violet. Violet has a huge variety of products from pop-up boxes to kitchen tissue roll for quick spills and kitchen prep.

    Why add to cart:

    ● Makes the gathering’s hosting and cleaning up easier

    ● Great for both outdoor and indoor events

    ● No extra cleaning equals more time to enjoy the celebration

    This Independence Day, shop Pakistani brands

    Buying local isn’t just about being patriotic, it’s about being practical and supporting local businesses. Pakistani products deliver quality and value that deserve to be in your cart all year round but especially on August 14. So whether you’re planning a celebration or just enjoying a chilled day at home, here’s your chance to shop smarter and shop local.

    And best of all, you can explore a huge variety of local Pakistani products at the tip of your fingers with Bazaar. With products across more than 30 categories to cover all your household and grocery needs, Bazaar has got you covered, delivering your order straight to your doorstep the very next day. No stress, no traffic, just convenience and celebration.

    Check out Bazaar now on mobile and web.

    This content is an advertisement by Bazaar and is not associated with or necessarily reflective of the views of Dawn.com or its editorial staff.

    Continue Reading

  • Researchers uncover the physics behind inaccuracies of cuff-based blood pressure readings

    Researchers uncover the physics behind inaccuracies of cuff-based blood pressure readings

    Researchers have found why common cuff-based blood pressure readings are inaccurate and how they might be improved, which could improve health outcomes for patients.

    High blood pressure, or hypertension, is the top risk factor for premature death, associated with heart disease, strokes and heart attacks. However, inaccuracies in the most common form of blood pressure measurement mean that as many as 30% of cases of high blood pressure could be missed.

    The researchers, from the University of Cambridge, built an experimental model that explained the physics behind these inaccuracies and provided a better understanding of the mechanics of cuff-based blood pressure readings.

    The researchers say that some straightforward changes, which don’t necessarily involve replacing standard cuff-based measurement, could lead to more accurate blood pressure readings and better results for patients. Their results are reported in the journal PNAS Nexus.

    Anyone who has ever had their blood pressure taken will be familiar with the cuff-based method. This type of measurement, also known as the auscultatory method, relies on inflating a cuff around the upper arm to the point where it cuts off blood flow to the lower arm, and then a clinician listens for tapping sounds in the arm through a stethoscope while the cuff is slowly deflated.

    Blood pressure is inferred from readings taken from a pressure gauge attached to the deflating cuff. Blood pressure is given as two separate numbers: a maximum (systolic) and a minimum (diastolic) pressure. A blood pressure reading of 120/80 is considered ‘ideal’.

    The auscultatory method is the gold standard, but it overestimates diastolic pressure, while systolic pressure is underestimated. We have a good understanding of why diastolic pressure is overestimated, but why systolic pressure is underestimated has been a bit of a mystery.”


    Kate Bassil, co-author from Cambridge’s Department of Engineering

    “Pretty much every clinician knows blood pressure readings are sometimes wrong, but no one could explain why they are being underestimated – there’s a real gap in understanding,” said co-author Professor Anurag Agarwal, also from Cambridge’s Department of Engineering.

    Previous non-clinical studies into measurement inaccuracy used rubber tubes that did not fully replicate how arteries collapse under cuff pressure, which masked the underestimation effect.

    The researchers built a simplified physical model to isolate and study the effects of downstream blood pressure – the blood pressure in the part of the arm below the cuff. When the cuff is inflated and blood flow to the lower arm is cut off, it creates a very low downstream pressure. By reproducing this condition in their experimental rig, they determined this pressure difference causes the artery to stay closed for longer while the cuff deflates, delaying the reopening and leading to an underestimation of blood pressure.

    This physical mechanism – the delayed reopening due to low downstream pressure – is the likely cause of underestimation, a previously unidentified factor. “We are currently not adjusting for this error when diagnosing or prescribing treatments, which has been estimated to lead to as many as 30% of cases of systolic hypertension being missed,” said Bassil.

    Instead of the rubber tubes used in earlier physical models of arteries, the Cambridge researchers used tubes that lay flat when deflated and fully close when the cuff pressure is inflated, the key condition for reproducing the low downstream pressure observed in the body.

    The researchers say that there is a range of potential solutions to this underestimation, which include raising the arm in advance of measurement, potentially producing a predictable downstream pressure and therefore predictable underestimation. This change doesn’t require new devices, just a modified protocol.

    “You might not even need new devices, just changing how the measurement is done could make it more accurate,” said Agarwal.

    However, if new devices for monitoring blood pressure are developed, they might ask for additional inputs which correlate with downstream pressure, to adjust what the ‘ideal’ readings might be for each individual. These may include age, BMI, or tissue characteristics.

    The researchers are hoping to secure funding for clinical trials to test their findings in patients, and are looking for industrial or research partners to help refine their calibration models and validate the effect in diverse populations. Collaboration with clinicians will also be essential to implement changes to clinical practice.

    The research was supported by the Engineering and Physical Sciences Research Council (EPSRC), part of UK Research and Innovation (UKRI).

    Source:

    Journal reference:

    Bassil, K. & Agarwal, A. (2025) Underestimation of systolic pressure in cuff-based blood pressure measurement. PNAS Nexus. doi.org/10.1093/pnasnexus/pgaf222

    Continue Reading