Citations

Clear

2117 citation(s) found for "Mayo, Steven".

The Survival Double Descent: Generalization Dynamics of Deep Neural Networks in Time-to-Event Analysis
Hart, Steven, Oberg, Ann L. , Hart, Steven N

Mayo Clinic

No container title available | March 4, 2026 | Volume: No volume available | Issue: No issue available | Pages: No page available

10.21203/rs.3.rs-8948458/v1 Get It

Abstract Recent work on double descent has challenged classical bias–variance tradeoffs, showing that test error can decrease, increase sharply near the interpolation threshold, and then decrease again as model capacity grows. This phenomenon has been documented in regression and classification, but its relevance to survival analysis remains unclear. Survival data are subject to censoring, which obscures true event times, and widely used models such as the Cox proportional hazards model are optimized via partial likelihoods that emphasize ranking rather than calibrated risk estimation. It is therefore unknown whether double descent occurs in this setting, how censoring influences its manifestation, or how it interacts with standard performance metrics.We investigate these questions using synthetic survival data generated from Weibull hazards with controlled censoring, allowing systematic variation of model capacity from under to over parameterized regimes. While we verified double descent occurs in survival models, calibration plateaus and decouples from discrimination, even under strong $\ell_2$ regularization. This decoupling arises because the Cox partial likelihood optimizes rankings rather than magnitudes, producing extreme risk scores that break the Breslow estimator used to estimate survival probabilities. These results highlight limitations of discrimination-based model selection in survival analysis and underscore the need for calibration-aware evaluation in high-capacity prognostic models.
Impact of Mavacamten Approval on Septal Reduction Therapy Rates in Hypertrophic Cardiomyopathy
Ibrahim, Ramzi, Wang, Winston , Pham, Hoang Nhat, Abdelnabi, Mahmoud, Lester, Steven J., Wilansky, Susan , Schaff, Hartzell V. , Geske, Jeffrey B., Arsanjani, Reza, Ayoub, Chadi , Geske, Jeffrey B. , Arsanjani, Reza , Ibrahim, Ramzi , Pham, Hoang Nhat , Ibrahim, Ramzi , Nhat, Hoang , Abdelnabi, Mahmoud , Geske, Jeffrey B , Arsanjani, Reza , Schaff, Hartzell

Mayo Clinic

Journal of the American Heart Association | March 4, 2026 | Volume: 15 | Issue: 6 | Pages: No page available

10.1161/jaha.125.045649 | PMID 41778598 Get It

Abstract PS3-06-11: A Novel Agent-Based AI System to Unlock Unstructured Data in Synoptic Reports for Advanced Population Health Analysis in Breast Cancer
Hart, S. N., Hart, Steven N

Mayo Clinic

Clinical Cancer Research | February 20, 2026 | Volume: 32 | Issue: 4_Supplement | Pages: PS3-06-11-PS3-06-11

10.1158/1557-3265.sabcs25-ps3-06-11 Get It

Abstract Background: While synoptic reports have standardized breast cancer data, a wealth of critical information remains locked in unstructured free-text, including nuanced biomarker details and procedural notes. This data is not easily computable, creating a significant barrier to understanding comprehensive population-level statistics and subtle clinical trends. Traditional data extraction methods are often insufficient to capture this complexity. We developed and validated a novel, flexible AI-powered system using an agent-based architecture to extract, normalize, and discretize this valuable unstructured data, enabling previously infeasible population-scale analytics. Methods: Our system employs a modular, agent-based framework where each "agent" is a specialized large language model (LLM) tasked with a specific function in a sequential workflow. The process begins with an initiator agent that parses the synoptic report. Subsequent agents then perform targeted tasks, such as identifying and normalizing tumor characteristics, procedural details, and biomarker information from free-text fields. A final agent aggregates this information into a structured, computable format. To rigorously test our approach, we first developed a synthetic data generator to create realistic, anonymized report snippets with known ground-truth values. We then evaluated the performance of several leading LLMs of various sizes (e.g., gemini, llama3.3, phi4, deepseek-r1, gemma3) on this synthetic dataset. Results: Our early analysis on simulated synoptic reports demonstrated high accuracy in data extraction and normalization. The gemini-1.5-pro-002 model achieved an overall accuracy of 96.9%, with a sensitivity of 99.3%, specificity of 94.3%, and an F1-score of 97.0%. The llama3.3:70b model also showed strong performance with a 98.7% accuracy, (F1: 98.8%, sensitivity: 99.8%, specifically: 97.6%). Gemma3:27b was the least performant model tested (F1: 66.2%), substantially worse than much smaller models (deepseek-r1:1.5b and phi4:14b). A detailed error analysis of false positives and negatives has allowed us to iteratively refine the agent logic. By the time of the symposium, we will have analyzed a large cohort of real, de-identified synoptic reports and will present comprehensive performance metrics and initial findings on population-level trends. Conclusion: This AI-powered, agent-based system offers a powerful and scalable solution to overcome the limitations of structured data in synoptic reports. By accurately extracting and normalizing previously inaccessible free-text information, our approach has the potential to provide clinicians and researchers with near real-time insights into population-level statistics and trends in breast cancer. This will facilitate a more profound understanding of disease patterns and ultimately support evidence-based advancements in patient care and clinical research. Citation Format: S. N. Hart. A Novel Agent-Based AI System to Unlock Unstructured Data in Synoptic Reports for Advanced Population Health Analysis in Breast Cancer [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS3-06-11.
Association of Trapezium Morphology With Primary Treatment Failure in Acute Traumatic Thumb Carpometacarpal Joint Dislocation
Nishimura, Ken , Moran, Steven L., Moran, Steven L , Moran, Steven L.

Mayo Clinic

HAND | February 20, 2026 | Volume: No volume available | Issue: No issue available | Pages: No page available

10.1177/15589447261422484 | PMID 41719050 Get It

Background: This study aimed to determine whether trapezium morphology, specifically trapezial inclination, is a risk factor for the failure of primary conservative or minimally invasive treatment for acute traumatic thumb carpometacarpal (CMC) joint dislocations in the relatively early period. We hypothesized that an elevated trapezial inclination would be associated with a higher rate of treatment failure. Methods: We retrospectively reviewed patients treated for acute thumb CMC joint dislocations between 1976 and 2025 with closed reduction and subsequent immobilization or percutaneous pinning. Patients were classified into “failure” (recurrent dislocation or subsequent symptomatic subluxation) or “non-failure” groups. Trapezial inclination, measured on oblique hand radiographs, was compared between the groups using the Mann-Whitney U test. Results: Fourteen patients met the inclusion criteria (7 failure, 7 non-failure). While demographics were not significantly different, the failure group demonstrated a significantly higher mean trapezial inclination (19.1° vs 8.7°; P = .0012). This significant difference persisted in a subgroup analysis that excluded associated Gedda type 3 Bennett fracture ( P = .0318) Conclusions: Elevated trapezial inclination may predict the failure of primary conservative or minimally invasive treatment in the relatively early period for acute thumb CMC joint dislocation. Our findings suggest that patients with this steep trapezium may be poor candidates for conservative or minimally invasive management and should be considered for primary definitive surgery.
Plasma Biomarkers of Senescence in Cholestatic Liver Disease: A Signature of Risk Stratification and Progression
O'Hara, Steven P. , Bogert, Pamela Tietz, Dickinson, Stephanie, Chen, Xiwei , Netto, Jair Machado Espindola, Kirkland, James L., Tchkonia, Tamara, Ashraf, Humza M. , Morgenthaler, Olivia , Ali, Ahmad H. , Gores, Gregory J., LaRusso, Nicholas, O’Hara, Steven P. , Bogert, Pamela S., Dickinson, Stephanie, Allison, David B. , Espíndola‐Netto, Jair Machado , Kirkland, James L. , Tchkonia, Tamar , Lazaridis, Konstantinos N. , Gores, Gregory J. , LaRusso, Nicholas F.

Mayo Clinic

Journal of Gastroenterology and Hepatology | February 14, 2026 | Volume: 41 | Issue: 3 | Pages: 1065-1076

10.1111/jgh.70279 | PMID 41689244 Get It

ABSTRACT Background and Aims Cellular senescence is a hallmark of several liver diseases, including primary sclerosing cholangitis (PSC) and primary biliary cholangitis (PBC). Senescent cholangiocytes exhibit a senescence‐associated secretory phenotype (SASP), characterized by profibroinflammatory mediator release. Current cost‐effective biomarkers predicting disease progression, particularly for PSC, are limited and often lack mechanistic relevance. We sought to define a plasma biomarker signature for PSC and PBC. Methods Plasma from early‐ and late‐stage PSC and PBC, alcoholic liver disease (ALD), inflammatory bowel disease (IBD) patients, and healthy controls was analyzed. Seventy‐one analytes were quantified using Luminex Multiplex Immunoassay or enzyme‐linked immunosorbent assay (ELISA). Principal component analysis (PCA) identified key patterns. Findings from the PSC cohort were then applied to additional cohorts. Results Second principal component (PC2) (17 analytes, 17.1% variability) best separated PSC from controls. ANOVA showed significant differences in PC2 between early PSC vs. controls ( p = 0.0001), late PSC vs. controls ( p < 0.0001), and early vs. late PSC ( p < 0.0001). PC2 analytes also distinguished early PBC vs. controls ( p < 0.0332), late PBC ( p < 0.0001), and ALD ( p < 0.0001), and early vs. late PBC ( p < 0.0001), but not IBD vs. controls ( p = 0.119). Logistic regression using PC2 demonstrated strong discrimination of early‐ and late‐stage PSC (AUC = 0.86) and control vs. early‐stage PSC (AUC = 0.83). Conclusion This is the first study to define a plasma SASP biomarker signature associated with cholestatic liver disease. These analytes track disease stage and represent both mechanistic indicators and potential clinical trial endpoints.
Colon-Pilot: A Generative AI Tool for Automated Colonoscopy Surveillance Recommendations and 2024 ACG/ASGE Quality Benchmarking
Garg, Sushil Kumar , Mau, Brayden , Hubers, Jeffery , Arce, Victor, Hooper, Nicole, Lindquist, Jennifer, Harper, Sarah B. , Thayer, Ann B., Mukherjee, Piyush, Loufek, Brenna, Melnick, Joe , Rost, Lauren, Robbins, Steven A. , Chaudhary, Rajeev, Harper, Sarah B , Loufek, Brenna T , Rost, Lauren M , Gomes, Rahul , Chaudhary, Rajeev , Hubers, Jeffrey , Arce, Victor, Mukherjee, Piyush, Rost, Lauren M. , Gomes, Rahul

Mayo Clinic

American Journal of Gastroenterology | February 10, 2026 | Volume: 121 | Issue: 4 | Pages: 964-973

10.14309/ajg.0000000000003946 | PMID 41665228 Get It

INTRODUCTION: High-quality colonoscopy requires accurate risk stratification for surveillance per the 2020 US Multi-Society Task Force guidelines and adherence to 2024 American College of Gastroenterology/American Society for Gastrointestinal Endoscopy quality benchmarks. Both are operationally challenging in clinical practice. We developed and validated Colon-Pilot, a large language model—powered clinical decision support system using GPT-4o to automate and standardize both functions. METHODS: The system was evaluated in 2 operational modes: (i) a human-in-the-loop clinical decision support validation of surveillance recommendations for 596 colonoscopies, comparing concordance with 2020 US Multi-Society Task Force guidelines against expert consensus, and (ii) an automated administrative audit applying Colon-Pilot to 42,632 colonoscopies across the Mayo Clinic Health System to calculate 2024 American College of Gastroenterology/American Society for Gastrointestinal Endoscopy priority quality indicators. Recommendations were autogenerated unless predefined safety criteria triggered manual review. RESULTS: Colon-Pilot issued recommendations for 522 of 596 cases (87.6%) and flagged 12.4% for manual review. For automated cases, guideline-concordant accuracy was 97.5% (Cohen κ = 0.970) versus 69.7% (κ = 0.781) for original endoscopist recommendations. Discordant artificial intelligence (AI) cases (n = 13) most often recommended longer-than-appropriate intervals (62%). Applied to the enterprise data set, Colon-Pilot calculated performance exceeding 2024 targets: adenoma detection rate 49.8% (≥35%), sessile serrated lesion detection rate 17.7% (≥6%), bowel preparation adequacy 91.8% (≥90%), and cecal intubation rate 97.4% (≥95%). DISCUSSION: Colon-Pilot demonstrated high fidelity in applying surveillance guidelines and automated quality benchmarking, outperforming unassisted endoscopists in guideline adherence. By combining safety protocols with large-scale automated reporting, it offers a scalable solution for improving both efficiency and quality in colorectal cancer prevention.
A Purified Platelet-Derived Exosome Product for Chronic Wound Healing: A Novel Therapeutic Strategy and Next-Generation Delivery Platform
Wan, Rou , Nishimura, Ken , Zhao, Chunfeng, Moran, Steven L. , Zhao, Chunfeng , Zhao, Chunfeng , Behfar, Atta

Mayo Clinic

Pharmaceutics | February 9, 2026 | Volume: 18 | Issue: 2 | Pages: 222

10.3390/pharmaceutics18020222 | PMID 41754964 Get It

Chronic wounds remain a major unmet clinical challenge, often failing to progress to normal healing due to persistent inflammation, impaired angiogenesis, and cellular senescence. Exosomes have recently been investigated as promising acellular therapeutics capable of restoring intercellular communication and promoting tissue regeneration. Among these, the Purified Exosome Product (PEP) represents a next-generation, platelet-derived exosome formulation manufactured under Good Manufacturing Practice (GMP) conditions with high purity, stability, and reproducibility. This review summarizes the current advances in exosome-based chronic wound therapeutics and PEP delivery systems and their translational potentials. Incorporation of PEP into bioengineered carriers such as fibrin sealant, collagen scaffolds, and hyaluronic acid (HA) hydrogels enables localized and sustained exosome release, significantly prolonging therapeutic effects and improving regenerative outcomes. Fibrin-based PEP delivery achieved complete wound closure and functional skin regeneration in animal models, while collagen and HA-based systems showed promising results for injectable and dermatologic applications. Beyond its intrinsic healing effects, PEP may also serve as a nanocarrier for other drugs, offering a future direction in chronic wound management.
Artificial Intelligence–Electrocardiography to Predict Incident Atrial Fibrillation and Clinical Outcomes in Kidney Transplant Recipients
Scalia, Isabel G., Farina, Juan M., Abdelnabi, Mahmoud H., Pietri, Milagros Pereyra, Ibrahim, Ramzi, Awad, Kamal, Abbas, Mohammed Tiseer , Ali, Nima Baba , Mahmoud, Ahmed K., Steidley, D. Eric, Masry, Hicham El , Sorajja, Dan , Scott, Luis R., Lee, Kwan, Johnson, D. W., Riad, Samy M. , Lester, Steven J., Jk, Oh, Ayoub, Chadi, Arsanjani, Reza, Farina, Juan M. , Ibrahim, Ramzi , Scalia, Isabel G. , Scalia, Isabel G , Farina, Juan M , Pereyra Pietri, Milagros, Pathangey, Girish , Ibrahim, Ramzi , Tiseer Abbas, Mohammed, Baba Ali, Nima , Alsidawi, Said , El Masry, Hicham , Sorajja, Dan , Scott, Luis R , Lee, Kwan S , Cho, Yeoungjee , Johnson, David W , Wadei, Hani M , Chao, Chieh-Ju , Mour, Girish K , Arsanjani, Reza , Abbas, Mohammed Tiseer , Pereyra, Milagros , Pathangey, Girish , Awad, Kamal , Mahmoud, Ahmed , Alsidawi, Said , Steidley, D. Eric , Scott, Luis R. , Cho, Yeoungjee , Wadei, Hani M. , Chao, Chieh‐Ju , Mour, Girish , Arsanjani, Reza

Mayo Clinic

Kidney360 | February 4, 2026 | Volume: 7 | Issue: 5 | Pages: 1128-1140

10.34067/kid.0000001063 | PMID 41637137 Get It

Key Points Atrial fibrillation after kidney transplantation is common and is associated with poor prognosis. Artificial intelligence applied to 12-lead electrocardiograms pretransplant accurately predicts the risk of new-onset atrial fibrillation in our cohort. High-risk artificial intelligence electrocardiography score was independently associated with increased risk of mortality and allograft failure. Background Incident atrial fibrillation (AF) is common after kidney transplantation (KTx) and is associated with worse clinical outcomes. Artificial intelligence electrocardiography (AI-ECG) algorithms have demonstrated efficacy in predicting risk of new-onset AF in the general population; however, their prognostic value in KTx recipients is relatively unknown. Methods Retrospective analysis was conducted on KTx recipients without AF, with at least one pretransplant ECG between 2011 and 2021 across three tertiary centers in the United States (Mayo Clinic sites in MN, AZ, and FL). A previously validated AI-ECG algorithm estimated the probability of incident AF for each patient. Based on AI-ECG probabilities, patients were categorized into high-risk and low-risk groups, with the optimal AI-ECG score cutoff determined. The incidence of new-onset AF, allograft failure, and mortality was compared between groups. Results Overall, 6246 patients (age 53.5±13.8 years; 58.9% male) were included. Pretransplant AI-ECG probability of AF ≥5% was the optimal cutoff for high risk of incident AF (sensitivity 72%, specificity 62%). High-risk scores were associated with true new-onset AF at 30 days (adjusted hazard ratio [aHR], 2.89; 95% confidence intervals [CI], 2.05 to 4.09; P < 0.001), 3 years (aHR, 2.54; 95% CI, 1.99 to 3.26, P < 0.001), and 5 years post-transplant (aHR, 2.48; 95% CI, 1.99 to 3.09, P < 0.001). High-risk AI-ECG scores were also associated with increased mortality (aHR, 1.56; 95% CI, 1.30 to 1.88, P < 0.001) and overall allograft failure (aHR, 1.50; 95% CI, 1.30 to 1.75, P < 0.001) through the 5-year follow-up. Conclusions This pretransplant AI-ECG parameter identified patients at increased risk of new-onset AF post-KTx and provided prognostic utility. Overall, this easy to obtain tool allows for risk stratification of patients who may benefit from closer monitoring, targeted risk factor modification, and early intervention.
Functional genomics identifies therapeutic options, biomarkers, and resistance mechanisms for high-grade gliomas
Lin, Wan-Hsin, Kosari, Farhad , Smadbeck, James B. , Barrett, Michael T. , Feathers, Ryan, Hall, Jamie, Sadeghian, Dorsay , Johnson, Sarah H. , Harris, Faye R. , Berry, Taylor, McCune, Alexa, Murphy, Stephen James, Kinsella, Lindsey, Ensign, Shannon P. Fortin, Yang, Lin , Emanuel, Angela, Jones, Leila A., Schaefer-Klein, Janet, Ida, Cristiane M., Sherman, Wendy, Porter, Alyx B. , Rosenfeld, Steven, Kizilbash, Sani H., Jaeckle, Kurt A. , Mrugala, Maciej M., Borad, Mitesh, Bendok, Bernard R., Quiñones-Hinojosa, A., Cheville, John C. , Vasmatzis, George , Anastasiadis, Panos Z. , Ensign, Shannon P. Fortin , Mansfield, Aaron S , Ensign, Shannon P. Fortin , Sotiriou, Sotiris , Haydu, Lauren E. , Moniz‐Garcia, Diogo , Ida, Cristiane M., Salomao, Marcela , Kizilbash, Sani H. , Mrugala, Maciej M., Mansfield, Aaron S. , Borad, Mitesh J. , Bendok, Bernard R. , Burns, Terry C. , Quiñones‐Hinojosa, Alfredo

Mayo Clinic

No container title available | February 3, 2026 | Volume: No volume available | Issue: No issue available | Pages: No page available

10.64898/2026.02.01.701806 | PMID 41676502 Get It

Abstract High-grade gliomas (HGGs) are aggressive tumors with poor outcomes and limited treatment options. Here, we combined genomic and transcriptomic tumor profiling with drug testing in a patient-derived 3-dimensional culture model to identify individualized treatments and predictive biomarkers. Activity of single agents targeting frequently dysregulated glioma pathways was relatively poor ex vivo and generally reflected historical patient data. However, compounds targeting PI3K, epigenetic, and survival/senescence signaling were effective in some cases. Drug sensitivity correlated with transcriptional rather than genomic features and suggested heterogeneity as a resistance mechanism. Bromodomain and extraterminal domain inhibition was particularly effective in tumors enriched in the mesenchymal transcriptional subtype, promoted proneural transition, and was overcome by upregulated PI3K signaling. Notably, combinations were largely effective, with 6 strategies exhibiting stronger efficacy than corresponding single agents in most cases (58-77%). This study identifies HGG vulnerabilities and associated biomarkers, resistance mechanisms, and effective combination strategies that warrant further clinical validation. Graphical Abstract
Endocannabinoids, perioperative pain, and acetaminophen in patients undergoing total knee arthroplasty: a prospective cohort study
Clendenen, Nathan , Clendenen, Anna M., McClain, Robert L., Wheeler, Garret, Christians, Uwe, Clendenen, Steven R. , Clendenen, Anna, McClain, Robert, Klawitter, Jost , Christians, Uwe , Klawitter, Jelena

Mayo Clinic

PAIN Reports | February 2, 2026 | Volume: 11 | Issue: 2 | Pages: No page available

10.1097/pr9.0000000000001369 | PMID 41635474 Get It

Abstract Introduction: Total knee arthroplasty (TKA) for knee osteoarthritis (OA) provides pain relief and functional improvement for many patients. The endocannabinoid system (ECS) regulates pain and inflammation through central and peripheral mechanisms and may be affected by acetaminophen. There is limited information on how surgery and perioperative care affect the ECS. Objective: We designed this study to examine the effect of surgery and acetaminophen on the ECS and determine whether perioperative endocannabinoids were associated with pain resolution after surgery. Methods: Forty adult patients with OA undergoing TKA were enrolled in a prospective observational cohort study and provided preoperative cerebrospinal fluid (CSF) and blood samples at 3 time points (before surgery, after intravenous acetaminophen, and 24 hours postoperative). Defense and Veterans Pain Rating Scale scores at rest and with activity were recorded in the preoperative holding area, in the operating room after intravenous acetaminophen before spinal anesthesia, and 24 to 48 hours postoperatively. Cerebrospinal fluid and plasma endocannabinoids were quantified using a validated liquid chromatography/mass spectrometry assay. We analyzed the data using linear regression models with sex, age, and body mass index as covariates. Results: Patients undergoing TKA have higher CSF and plasma concentrations of N-acylethanolamines anandamide and its congeners compared with control subjects. Patients with higher pain scores had lower CSF anandamide levels before and after surgery and higher levels of 2-arachidonoylglycerol before surgery but not after. Conclusion: Chronic pain, acetaminophen, and surgery are associated with endogenous endocannabinoid levels and may present an opportunity for developing targeted nonopioid perioperative analgesia after TKA.