Citations

Abstract Rationale Quantitative CT using multiple instance learning (MIL)-usual interstitial pneumonia (UIP) classifier and data-driven textural analysis (DTA) provides objective and reproducible measurements of disease pattern and extent, respectively. However, the understanding of their individual or combined relationships with outcomes in fibrotic hypersensitivity pneumonitis (fHP) remains limited. Objectives To determine whether visual CT patterns of fHP or UIP, a MIL-UIP classifier, baseline and changes in DTA over 12-months and DTA at 12 months are associated with progression-free survival (PFS) in fHP. Methods The primary cohort included 134 participants with fHP. Associations between CT variables and subsequent lung function change, and PFS were assessed using linear mixed models and Cox hazards models. Results were validated in a combined dataset from three independent populations (N = 216). Results Visual patterns of fHP or UIP were not associated with PFS. A baseline MIL-UIP classification and DTA were significantly associated with 24-month PFS. However, the MIL-UIP did not predict PFS after adjusting for DTA fibrosis extent. Over 12 months, a 1-point increase in DTA across and within patients resulted in a significant decline in FVC% and DLCO%. A 12-month change in DTA (HR = 1.07, 95%CI 1.01-1.13) and DTA at 12-months (HR = 1.03, 95%CI 1.01-1.05) were associated with subsequent 12-month PFS. Similar estimates were observed in the validation cohort. Conclusions In fHP, the objectively measured extent of lung fibrosis is more prognostic than either a UIP classifier or visual CT patterns of fHP or UIP. Baseline DTA, DTA at 12 months, and its change over 12-months are associated with PFS.

Abstract M.D.-Ph.D. programs in the United States have traditionally followed a “2- n -2” curricular model, with the graduate phase occurring between the pre-clerkship and clerkship portions of medical training. While well established, this format can limit trainee autonomy in shaping their physician-scientist training trajectories. In response, some programs have introduced a “3- n -1” model, allowing students to complete clerkships before beginning Ph.D. training. Our institution implemented multiple flexible curricula in 2017. Understanding why trainees choose one pathway is important as programs consider implementing more adaptable training structures. To investigate these factors, we surveyed M.D.-Ph.D. students at our institution, which offers multiple flexible curricular alternatives, about contributions to their curricular decisions. Responses supported that trainees weigh considerations across medical education, scientific development, and integrated physician-scientist training domains. Although the traditional pathway was a popular option, most students pursued one of the flexible pathways. Our findings suggest that introducing flexibility in the timing of undergraduate medical education and graduate training can support diverse educational, logistical, and personal priorities while maintaining the rigor of physician-scientist training. Offering multiple pathways empowers trainees to design trajectories that best fit their needs. Continued longitudinal studies are needed to assess the long-term impacts of flexible curricula on physician-scientist career outcomes.

Objective To examine sex differences in the relationship between comorbid anemia and heart failure (HF) with 13-year all-cause mortality in Mexican American older adults aged 75 years and older. Methods Participants ( N = 1,615) aged ≥75 years were from the Hispanic Established Population for the Epidemiologic Study of the Elderly (2004/05–2016). Participants were categorized by anemia status, HF status, and sex. Cox proportional hazards regression models estimated hazard ratios (HRs) and 95% Confidence Intervals (CIs) for all-cause mortality. Results HF only was associated with higher mortality in males (HR = 1.41, 95% CI = 1.12–1.78; p = 0.0039) and females (HR = 1.52, 95% CI = 1.24–1.85; p < 0.0001). Anemia only (HR = 1.54, 95% CI = 1.06–2.23; p = 0.0248) and comorbid anemia and HF (HR = 1.74; 95% CI = 1.00–3.03; p = 0.0484) were associated with higher mortality only in males. Conclusion Anemia and HF were associated with greater mortality risk among older Mexican American males than females.

Antiretroviral therapy has transformed treatment of HIV infection from a fatal disease into a chronic condition, yet opportunistic infections remain prevalent in patients with declining CD4 cell counts, requiring a pattern-based radiologic approach and vigilant clinical assessment to differentiate infections, malignancies, and immune reconstitution inflammatory syndrome.

Abstract Background Previous studies evaluated large language model (LLM) performance on the American Society of Nuclear Cardiology (ASNC) Board Preparation Exam. Without domain-specific context, the best model (GPT-4o) achieved 63.1%, below the estimated 65% passing threshold and the 78% mean score of human fellows-in-training (FITs). Providing textbook context improved GPT-4o to 73.8% on text-only questions, but still fell short of human trainees. Whether next-generation LLMs with retrieval-augmented generation (RAG) can exceed this gap is unknown. Methods Claude Opus 4.7 and GPT-5.5 were administered all 168 questions (141 text-only, 27 image-based) from the 2023 ASNC Board Preparation Exam across 5 iterations each, using RAG with a nuclear cardiology textbook, companion atlas, and ASNC clinical guidelines. Claude used local FAISS-based semantic retrieval; GPT-5.5 used Azure’s cloud-hosted vector store. Performance was compared to prior LLM results and 13 human FITs. Results Across 5 iterations, Claude Opus 4.7 achieved a mean accuracy of 86.3% ± 1.4% (text 88.8%, image 73.3%). GPT-5.5 achieved 86.7% ± 2.2% (text 88.5%, image 77.0%) but refused a mean of 12.2 questions (7.3%) per iteration due to safety filters. Both models surpassed the human FIT mean (78.0%) and the estimated passing threshold. Compared to GPT-4o without context (63.1%), this represents a 23-percentage-point improvement in 18 months. Conclusion Next-generation LLMs with RAG now surpass average human trainee performance on nuclear cardiology board preparation questions, suggesting significant potential as educational tools and knowledge-reference aids in cardiovascular imaging. Condensed Abstract Across 5 iterations each, Claude Opus 4.7 and GPT-5.5 with retrieval-augmented generation achieved mean accuracies of 86.3% and 86.7% on the 2023 ASNC Board Preparation Exam (168 questions), both surpassing the mean human fellow-in-training score of 78%. GPT-5.5 refused a mean of 12.2 questions (7.3%) per iteration due to safety filters. These results represent a 23-percentage-point improvement over the best prior LLM without context (63.1%), demonstrating that RAG-enhanced LLMs have reached human-level proficiency in nuclear cardiology knowledge. Graphical Abstract Overview of the three-study research arc evaluating LLM performance on the 2023 ASNC Board Preparation Exam. Study 1 (2024) tested four LLMs without context (best: GPT-4o, 63.1%). Study 2 (2025) added textbook context to GPT-4o (73.8%). Study 3 (2026, current) evaluated Claude Opus 4.7 and GPT-5.5 with retrieval-augmented generation across 5 iterations each (mean 86.3% and 86.7%, respectively), both surpassing the human fellow-in-training mean of 78%. Right panel shows the performance scale with key thresholds.

BACKGROUND AND OBJECTIVES: Craniopharyngiomas are complex intracranial tumors that pose significant challenges in surgical management. We aimed to develop risk-stratified surgical outcome benchmarks for adult patients undergoing endoscopic craniopharyngioma resection by using the Registry of Adenomas of the Pituitary and Related Disorders database to facilitate comparisons across institutions and guide clinical decision-making. METHODS: A retrospective cohort of adult patients (18 years and older) with craniopharyngioma undergoing a first-time endoscopic transnasal surgery across 14 academic centers in the United States was analyzed using data collected from 2011 to 2024. Patients were stratified into low-risk and high-risk cohorts based on criteria including age, body mass index, selected tumor characteristics (size, infundibular relationship, intraventricular extension), and previous radiation. Surgical outcomes were analyzed, and benchmarking ranges were defined using the established methodology. RESULTS: The overall cohort included 318 patients with a mean age of 50 years. Upon benchmarking analysis (50th-75th percentile performance), low-risk patients (n = 136) had a median length of stay of 4.0 to 6.0 days, 90-day readmission rates of 15.5% to 22.2%, skilled nursing facility discharge rates of <1.0%, postoperative cerebrospinal fluid leak incidence of <1.0% to 14.3%, gross total resection rates of 50.0% to 60.8%, new postoperative arginine vasopressin deficiency rates of 33.3% to 50.0%, and tumor progression/recurrence rates of <1.0% to 18.3% over a median follow-up period of 27.0 (IQR 10.5-56.5) months. Benchmarking ranges were proposed for each outcome to facilitate cross-institutional comparisons. CONCLUSION: Establishing surgical outcome benchmarks for adult patients undergoing craniopharyngioma resection using data from the Registry of Adenomas of the Pituitary and Related Disorders provides a framework for assessing surgical performance and guiding clinical decision-making. Benchmark ranges can inform institutional quality improvement efforts, perioperative risk stratification, and patient counseling regarding expected postoperative outcomes.

Purpose: The study compares refractive outcomes of secondary intraocular lens (SIOL) surgery performed by vitreoretinal, anterior segment, and both surgeons. Methods: Eyes included in this study underwent SIOL surgery from 2012 to 2024 and had a final postoperative visual acuity of 20/50 or better. Regression analysis compared refractive outcomes, and chi-squared analysis compared complications between subspecialty of SIOL surgeons. Results: A total of 94, 104, and 80 eyes were included in the retina surgeon, anterior segment surgeon, and combo surgeon cohorts, respectively. There was no significant difference in error between target and final postoperative spherical equivalence between retina and combo ( P = 0.079), retina and anterior segment ( P = 0.135), or anterior segment and combo cohorts ( P = 0.567). There was no significant difference in number of eyes with final spherical equivalence within 1 diopter of target spherical equivalence between retina and combo ( P = 0.136), retina and anterior segment ( P = 0.061), or anterior segment and combo cohorts ( P = 0.783). The retina cohort had higher postoperative corneal edema than the combo ( P = 0.017) and the anterior segment cohorts ( P = 0.035). The combo cohort had more postoperative retinal detachment than the anterior segment cohort ( P = 0.046). Conclusion: In eyes with good final vision, retina and anterior segment surgeons have similar refractive outcomes after SIOL surgery.