Citations

Abstract Large language models (LLMs) and vision‐language models represent a fundamentally different category of artificial intelligence (AI) compared to prior image analysis approaches in digital pathology, which have largely been based on convolutional neural network architectures. This review from the American Society of Cytopathology Clinical Practice Committee examines the current evidence for LLM and vision‐language model applications in cytopathology, including structured reporting, diagnostic assistance, quality control, education, and workflow integration. The distinction between applications with preliminary evidence and those that remain hypothetical is described. A detailed assessment of the challenges that must be addressed before clinical deployment, including hallucination risk, limited explainability, bias, data privacy, validation gaps, and infrastructure barriers is discussed. A review of the regulatory landscape in the United States and European Union as it applies to AI‐enabled software as a medical device is provided. Recommendations addressing cytopathology‐specific benchmarks, multi‐institutional validation, transparent governance, and incremental deployment beginning with low‐risk applications are suggested. In the current environment, LLMs have the potential to augment cytopathology practice, but responsible adoption requires rigorous validation and sustained collaboration among cytopathologists, AI researchers, and regulatory bodies.

ABSTRACT Objectives To investigate pharmacological treatment patterns in individuals with bipolar disorder (BD) with and without comorbid substance use disorder (SUD) and anxiety disorder (AX), we leveraged the Global Bipolar Cohort to analyze cross‐regional practices across North America, Europe, and the Pacific. Methods Fourteen cohorts contributed aggregate data on pharmacotherapy, demographics, diagnostic subtypes, and comorbidities. Proportional meta‐analyses using generalized linear mixed models were conducted to examine prescription trends and identify clinical differences. Results The sample ( N = 11,521) was 60% female and 84% Caucasian. Participants were categorized into four mutually exclusive groups based on comorbidity status: those with comorbid AX only, comorbid SUD only, both AX and SUD, or neither. The AX+SUD subgroup showed higher rates of attention‐deficit/hyperactivity disorder (ADHD), post‐traumatic stress disorder (PTSD), rapid cycling, obesity, and unemployment, reflecting a more severe clinical profile. Regional variations were notable: North American cohorts reported higher prevalence of AX and SUD than European and Pacific cohorts. Antidepressants use for AX were more common in Europe and the Pacific, while North American prescribing patterns were more variable. Benzodiazepine use was high among individuals with SUD across all regions. Lithium and first‐generation antipsychotic prescriptions varied, with higher rates observed in Europe. Conclusions Findings underscore the heterogeneity of BD and the influence of comorbid AX and SUD on illness burden and treatment. Regional prescribing variations underscore the need for context‐specific guidelines. Gaps in data on medication‐assisted treatment for SUD point to areas for future research. These insights can support more individualized and effective care for complex BD presentations.

Summary Background Body composition is recognized as a major determinant of health outcomes, but its multidimensional nature makes clinical adoption challenging. We sought to develop and validate a body composition index (BCI) for all-cause mortality risk assessment, integrating variables of six body composition tissues. Methods We analyzed 28509 consecutive patients undergoing myocardial perfusion imaging with routine low-dose chest CT attenuation correction (CTAC) scans acquired during myocardial perfusion imaging (MPI) at 12 centers across four countries. An artificial intelligence-based BCI was developed in a cohort of 15037 patients’ CTACs by integrating the CT-derived metrics of bone, skeletal muscle, and four adipose tissue compartments, coronary artery calcium score, and basic demographic variables (age, sex, BMI). The performance of BCI for mortality prediction was validated in an internal cohort of 6444 patients and an external cohort of 7028 patients by prognosis, calibration, net benefit, and explainability. Model-based simulation of tissue metrics modification was performed to evaluate estimated mortality risk reduction. Findings During a median of 3.5 (IQR [1.9, 5.1]) years, 4697 (16%) patients died. In the external testing cohort, the BCI demonstrated excellent discrimination for mortality (area under receiver operating characteristic curve 0.78 (95% CI [0.76, 0.79]) and Harrell’s concordance index 0.75 [0.73, 0.76]), calibration, and net benefit overall and across pre-specified subgroups stratified by patient characteristics and imaging protocols. Visceral adipose tissue attenuation was the most influential body composition measure, followed by skeletal muscle volume. Simulated improvement in body composition was associated with significant mortality risk reduction. Interpretation An index combining six body composition measures obtained opportunistically from routine chest CT provides robust mortality risk stratification. By converting complex body composition information into a single interpretable score, the BCI can facilitate clinical implementation of opportunistic CT biomarkers and guide individualized preventive strategies. Funding National Institute of Health Research in context Evidence before this study We searched PubMed and Google Scholar on July 7, 2026 for English language studies using the terms (“mortality prediction” OR “death prediction”) AND (“integrated body composition analysis” OR “body composition metrics integration”) AND (“predictive model” OR “model development”), and identified 41 publications describing predictive models incorporating body composition metrics. Recent advances in artificial intelligence (AI) have enabled increasingly granular quantification of body composition from CT, with multiple tissues and tissue-specific metrics showing independent prognostic value for all-cause mortality (ACM) prediction. However, reported associations of body composition metrics with mortality risk were highly inconsistent or even contradictory across studies. This heterogeneity likely reflected the differences in study design or the inherent complexity of body composition analysis, such as the interaction among variables derived from multiple tissues. Thus, it is difficult to interpret body composition results reliably in clinical practice. We hypothesized that integrating multiple body composition metrics could achieve comprehensive and reliable prediction of mortality across patient populations. Added value of this study In this large international, multicenter longitudinal cohort study of 28509 consecutive patients with suspected or known coronary artery disease (CAD) undergoing myocardial perfusion imaging across 12 sites in four countries, we developed and externally validated an artificial intelligence-based body composition index (BCI) for all-cause mortality prediction. The BCI integrated chest CT-derived metrics of six body composition tissue compartments – bone, skeletal muscle (SM), subcutaneous adipose tissue (SAT), intramuscular adipose tissue (IMAT), visceral adipose tissue (VAT), and epicardial adipose tissue (EAT), together with demographics (age, sex, BMI) and CT-derived coronary artery calcium (CAC) score, all derived automatically from routine low-dose attenuation correction CT scans without additional imaging or laboratory tests. In an external validation cohort of 7028 patients, the BCI achieved robust discrimination for mortality prediction, substantially outperforming a baseline model using demographics and calcium score alone, and exceeded the discriminative performance of prior body composition models. Performance was consistent across subgroups defined by sex, age, body mass index, cardiometabolic risk factors, imaging modality, and acquisition protocol, supporting generalizability across clinical settings. The BCI demonstrated excellent calibration and net clinical benefit. Explainability analysis identified visceral adipose tissue attenuation and skeletal muscle volume as the dominant contributors to mortality risk, offering tissue-specific therapeutic targets that complement generic weight-based metrics. Model-based simulations suggested that improvements in these specific compartments could reduce estimated mortality risk more effectively than equivalent changes in BMI alone, pointing toward precision therapeutic targets beyond weight reduction. Our study improved over previous literatures on body composition-based predictive models in one or more of the following ways: (1) we considered six types of body composition tissues while existing studies considered only three or less tissues, thus significantly higher granularity and improved ability to capture refined tissue interactions; (2) our model was rigorously validated in external sites, showing cross-site generalizability while most previous work lacked external validation; (3) our model achieved significant prediction performance improvement over previous body composition predictive models, offering better clinical outcomes prediction and facilitating the suitability for widespread clinical adoption; (4) our study was conducted with a large multi-center cohort across multiple countries, achieving statistical power and demonstrating reliability towards patients characteristics and imaging acquisition protocols while previous studies involved small single-center cohorts; (5) our model required only CT scans and basic demographics which were widely available in clinical routine while the existing models required dedicated laboratory tests or specialized imaging modalities, enhancing the applicability in routine clinical practice. Implications of all the available evidence BCI offers an explainable and clinically deployable tool for improved prediction of mortality risk from existing chest CT scans. Its use could inform updates to existing risk models, guide precision preventive strategies, and support development of therapeutics targeting specific body composition phenotypes.

Importance Pyoderma gangrenosum (PG) is a rare neutrophilic dermatosis characterized by sterile ulcerative skin lesions. Multiple comorbidities and clinical features have been associated with PG, but no standardized guidelines exist for classifying PG phenotypes. Objective To develop an expert-established classification framework for PG phenotypes to inform treatment guidelines and future research endeavors. Evidence Review In this modified Delphi consensus study that included 23 board-certified dermatologists and Medical Dermatology Society members with expertise in PG, panelists completed 5 rounds of anonymous, iterative online surveys that were administered from December 2023 through July 2025. Before the beginning of the consensus exercise, a literature review was performed by nonvoting researchers to summarize existing data on PG clinical associations and comorbidities. A PubMed search from inception to September 2023 identified observational studies, narrative reviews, systematic reviews, and meta-analyses describing PG clinical associations and comorbidities that were published in English. Experts indicated their agreement with proposed PG phenotypes and disease modifiers with a consensus threshold of 70% or greater. Anonymous comments and aggregated results were presented in each subsequent round. In the final round, a framework of PG phenotypes and disease modifiers was proposed for agreement. Findings Twenty-three board-certified dermatologists and Medical Dermatology Society members with expertise in PG completed 5 rounds of iterative surveys. Consensus was reached on the final set of PG phenotypes and disease modifiers, with 83% of experts in agreement. PG phenotypes were overall classified into 2 major groups: PG with autoinflammatory syndromes and nonsyndromic PG. Nonsyndromic PG included 4 phenotypes: inflammatory bowel disease–associated PG, PG in association with hematologic cancers and blood dyscrasias, drug-induced PG, and other (including idiopathic) PG. Disease modifiers of PG phenotype presentations included involvement of special sites (head/neck, genitals, or peristomal skin) and extracutaneous manifestations. Conclusions and Relevance This expert-established, descriptive framework provides a standardized classification system for distinct PG phenotypes and its modifiers. This nomenclature may inform upcoming clinical guidelines and allow for consistency in reporting epidemiological research and outcomes among patients with PG.

OBJECTIVE Cognard type V dural arteriovenous fistulas (dAVFs), defined by spinal perimedullary venous drainage, represent a rare and aggressive neurovascular pathology. Comprehensive multicenter data remain scarce. The authors present the largest multicenter analysis to date, characterizing presentation, radiology, treatment, complications, and functional outcomes of Cognard type V dAVFs. METHODS The authors retrospectively analyzed patients with angiographically confirmed Cognard type V dAVFs from the Consortium for Dural Arteriovenous Fistula Outcomes Research (CONDOR), spanning 16 academic centers. Patient demographics, imaging features, treatment strategies, procedural complications, and modified Rankin Scale (mRS) scores were recorded. RESULTS Among 1077 patients with dAVFs, 37 (3%) had Cognard type V lesions. The mean patient age was 57 years, and 59.5% of patients were male. Most presented with nonhemorrhagic neurological deficits (68%), while 14% had hemorrhage. The most common location was the foramen magnum (32%). Embolization was the primary treatment in 67% of cases, complete obliteration was achieved in 71% of cases. Upfront microsurgery was performed in 22% of patients (obliteration rate 88%), and radiosurgery in 14% (success rate 40%). Salvage treatment further improved obliteration outcomes. Temporary complications occurred in 8% of patients (3/37): 2 with transient neurological deficits postembolization and 1 with transient hydrocephalus after surgery. Permanent complications were observed in 3% (1/37) due to treatment-related hemorrhage. At the last follow-up (mean 2.4 years), 69% (24/35) of patients maintained or improved to an mRS score ≤ 2. No deaths occurred. Functional improvement was seen in 29% (10/35), while mRS decline was in 40% of patients (14/35), all unrelated to the dAVF itself. CONCLUSIONS Cognard type V dAVFs carry a high risk of neurological morbidity. However, favorable functional outcomes are achievable with timely, multimodal intervention. Complication rates remain low, with most being transient. These data support aggressive but tailored management to prevent permanent disability in this high-risk population.

Objective: To determine whether sleep apnea–specific hypoxic burden (SASHB) is associated with self-reported hypertension beyond conventional obstructive sleep apnea metrics in rail workers. Methods: Cross-sectional study of 84 Argentine rail workers undergoing home sleep apnea testing. Associations between SASHB and hypertension were evaluated using adjusted logistic regression. Drowsy driving was examined as an unadjusted exploratory secondary outcome. Results: Higher SASHB was independently associated with hypertension in the fully adjusted model (OR per SD 2.42; 95% CI 1.24–5.29; p=0.015). Higher SASHB was also associated with more severe self-reported drowsy driving in unadjusted exploratory ordinal analyses (OR per SD 1.78; 95% CI 1.09–3.28; p=0.035). Conclusions: SASHB may provide complementary information beyond conventional OSA metrics for cardiovascular risk assessment in occupational populations.

Background With improvements in in vitro culture techniques, there has been a steady shift in practice to transfer in vitro fertilisation (IVF) embryos at the blastocyst-stage to increase pregnancy rates per embryo transfer (ET) and reduce the risk of multiple pregnancy. In patients with fewer embryos (≤5 zygotes), fresh cleavage-stage ET is still commonly performed to reduce the risk of cycle cancellation if no embryo survives to the blastocyst-stage in vitro. However, evidence for improved outcomes with cleavage-stage embryo transfer in this subgroup of patients is lacking. We will compare blastocyst-stage to cleavage-stage ET with regard to cumulative live birth rates (CLBR) per oocyte retrieval and other patient-centred cycle outcomes in patients with ≤5 zygotes. Methods and analysis Multicentre, randomised pragmatic comparative effectiveness trial with superiority design comparing blastocyst-stage to cleavage-stage ET in patients with ≤5 zygotes. This trial will enrol and randomise 1126 women aged 18–44 undergoing IVF to either fresh cleavage-stage or blastocyst-stage ET at eleven IVF centres. Surplus embryos will be vitrified at the blastocyst-stage. The primary outcome is the CLBR per oocyte retrieval. Secondary outcomes include: multiple pregnancy, pregnancy loss <20 weeks and cycle cancellation. Exploratory outcomes include: live birth per fresh ET, ectopic pregnancy, time to pregnancy, stillbirth, preterm birth, financial burden, patient distress and facilitators and barriers for patient involvement and implementation. All primary analyses will be conducted according to the intent-to-treat principle. We also will report per-protocol analyses as a sensitivity analysis. A multidisciplinary stakeholder team, including a 5-patient stakeholder panel, will participate in study design, conduct and oversight. Ethics and dissemination The Institutional Review Board at Beth Israel Deaconess Medical Center (BIDMC) provides central regulatory oversight. Results will inform timing of embryo transfer in patients with ≤5 zygotes and will be disseminated through academic publications, meeting presentations and communications to advocacy groups and patient participants. Trial registration number NCT06746129 .