In press
AI-Assisted Hypothesis Generation for Hospital Quality Improvement
Vossler*, Feng*, Sivaraman, Gallo, Kanzaria, Freiser, Ross, Ou, Marks, Ehrlich, Peabody and Zier
npj Health Systems, In press
[paper]
AppliedPROSPECTLLMs & generative AIHuman-AI interaction
Preprints
Expert Evaluation of Clinical AI Tools on Real Point-of-Care Clinical Queries
Feng, Patel, Heagerty, Mai, Sivaraman, Vossler, Ouyang and Jena
[arXiv][code][data]
AppliedEvaluation & auditingLLMs & generative AI
Expert-guided g-computation with large language models for estimating causal effects on timings: Applications to hospital quality improvement
Vossler, Ouyang, Guo, Huang, Shojaie, Zier, Xia and Feng
[arXiv]
MethodsPROSPECTCausal inferenceLLMs & generative AI
``I know where to look,’’ but does the LLM? Charting the gaps between clinical expert needs and unstructured data abstraction tools
Sivaraman, Turnham, Bonano, Aresh, Dhanasekaran, Guo, Kubendran, Lin, Louie, Olazo, Shen, Vasudevan, Wong, Alsentzer, Fries, Odisho, Gordan, Feng* and Hong*
[arXiv][code]
AppliedLLMs & generative AIHuman-AI interaction
Learning to detect shifts in unstructured data streams via monitoring-by-betting
Zhou, Ouyang, Xia, Subbaswamy, Prinster, Saria, Garcia-Agundez, Pollet, Yazdany, Vossler* and Feng*
[arXiv][code]
MethodsMonitoring & updatingLLMs & generative AI
A framework to monitor editing of Artificial Intelligence-generated medical documentation
Garcia-Agundez, Zhou, Pourian, Blebea, Daneshpajouhnejad, Dente, Shi, Pollet, Xia, Shi, Thombley, Fenton, Murray, Adler-Milstein, Schmajuk, Feng* and Yazdany*
[medRxiv]
AppliedMonitoring & updatingLLMs & generative AI
Indirect influences of artificial intelligence advice on complex sepsis treatment decisions
Sivaraman, Levin, Mason, Talisa, Saria, Feng, King, Kahn and Perer
[arXiv]
AppliedHuman-AI interaction
2026
Adaptive auditing of AI systems with anytime-valid guarantees
Zhou*, Vossler*, Sivaraman, Mai and Feng
Adv. Neural Inf. Process. Syst., 2026
[paper][code][demo] Stellar Abstract Award at STAI-X
MethodsEvaluation & auditingLLMs & generative AI
When the domain expert has no time and the LLM developer has no clinical expertise: Real-world lessons from LLM co-design in a safety-net hospital
Kothari, Vossler, Digitale, Forouzannia, Rosenberg, Lee, Bryant, Molina, Marks, Zier and Feng
Proc. Conf. AAAI Artif. Intell., 2026
[paper][code] Oral presentation
AppliedPROSPECTLLMs & generative AIHuman-AI interaction
LLMs Judging LLMs: A Simplex Perspective
Vossler, Xia, Mai, Subbaswamy and Feng
International Conference on Artificial Intelligence and Statistics (AISTATS), 2026
[paper][code][poster]
MethodsEvaluation & auditingLLMs & generative AI
Scaling Medical Device Regulatory Science using Large Language Models
Li, He, Subbaswamy, Vossler, Gossmann, Singh and Feng
NPJ Digit. Med., 2026
[paper][code]
AppliedRegulatory scienceLLMs & generative AI
Natural language processing to identify documented pain preceding radiotherapy for bone metastases
Chen, Feng, Friesner, Chang, Kenny, Elia, Ni, Feng, Rabow, Braunstein, Boreta and Hong
JNCI Cancer Spectrum, 2026
[paper]
AppliedLLMs & generative AI
Patient perspectives about deployment of artificial intelligence decision support tools in a safety-net healthcare system
Phelps, Vossler, Kothari, Steineman, Goldman, Gasga, Ehrlich, Kanzaria, Adler-Milstein, Feng, Nong and Zier
JAMIA Open, 2026
[paper]
AppliedPROSPECTHuman-AI interaction
More than ``means to an end’’: Supporting reasoning with transparently designed AI data science processes
Sivaraman, Vossler, Perer, Hong and Feng
ACM CHI 2026 Workshop on Tools for Thought, 2026
[paper]
PerspectiveHuman-AI interaction
Clinical trials for continuously monitored and updated AI systems
van Amsterdam*, Oberst*, Feng*, Wiens, Tang, Joshi, Ranganath, Sendak, Shalit, Vogt, Beaulieu-Jones, Mamdani, Kent, Heagerty, Fleming and Goldenberg
Nat. Med., 2026
[paper]
PerspectiveMonitoring & updatingRegulatory science
Human-AI co-design for clinical prediction models
Feng*, Kothari*, Vossler, Bishara, Zier, Addo, Kornblith, Tan and Singh
npj Digital Medicine, 2026
[paper][code][poster]
MethodsPROSPECTInterpretabilityLLMs & generative AIHuman-AI interaction
STAVAT: Stratified Anytime-Valid Atom Sampling for LLM Tier Placement
Vossler, Xia and Feng
Statistics and Trustworthy AI for Cross (X)-Domain Acceleration (STAI-X), 2026
[paper] Paper Award
MethodsEvaluation & auditingLLMs & generative AI
Assessing acuity in pediatric emergency department triage: performance of a large language model
Narang, Addo, Williams, Firnberg, Feng and Kornblith
Npj Health Syst., 2026
[paper]
AppliedEvaluation & auditingLLMs & generative AI
Building a global research network for Fair, Accountable, Interpretable, and Responsible AI in Emergency Care: Protocol for a FAIR-EC study
Hong, Liew, Yu, Barry, Blewer, Buckland, Cai, Cha, Chakraborty, Chen, Cheng, Chong, Djärv, Earnest, Engelhard, Fan, Feng, Feng, Fu, Goh, Goldstein, Gronsbell, Ho, Ho, Iwami, Joiner, Kornblith, Li, Lim, Liu, Liu, Lu, Luo, Ng, Ning, Okada, Park, Park, Razzak, Shen, Siddiqui, Steel, Tan, Teixayavong, Vakulenko-Lagun, Vissoci, Waligora, Wang, Wang, Wang, Wong, Xie, Yang, Zhang, Zhou, Zhou, Zhu, Neumar, Page, Vaughan, Ong and Liu
JMIR Res. Protoc., 2026
[paper]
AppliedEvaluation & auditing
A proactive digital valve care pathway for management of severe aortic stenosis: The DETECT AS II study
Forouzannia, Hempstead, Steineman, Duan, Vossler, Feng, Zhou, Zier and Elmariah
JAMA Cardiol., 2026
[paper]
AppliedClinical prediction models
Clinical AI in 2025–2026 — six themes from the SAIL year in review
Kim, Elia, Li, Healey, Erdman, Chandak, Argaw, Li, Black, Marwaha, Morales, Taylor, Kvedar, Brodeur, Beaulieu-Jones, Alsentzer, Hong and Feng
NEJM AI, 2026
[paper]
Review
Adjusted trend analysis in the adoption and outcomes of minimally invasive pancreaticoduodenectomy approaches: over a decade of contemporary U.S. data
Foroutani, Chang, Kannan, Ganjouei, Wang, Feng, Thornblade, Hirose, Maker, Nakakura, Corvera, Alseidi and Adam
Surg. Endosc., 2026
[paper]
AppliedClinical outcomes research
2025
Realistic CDSS drug dosing with end-to-end recurrent Q-learning for dual vasopressor control
Zou, Feng, Kalimouttou, Zhang, Seymour and Pirracchio
NeurIPS 2025 Workshop on Learning from Time Series for Health, 2025
[paper]
MethodsCausal inference
Not all clinical AI monitoring systems are created equal: Review and recommendations
Feng, Xia, Singh and Pirracchio
NEJM AI, 2025
[paper]
ReviewMonitoring & updating
Reducing readmissions in the safety net through AI and automation
Bennett, Feng, Goldman, Kothari, Gottlieb, Durstenfeld, Marks, Ehrlich, Davis and Zier
Am. J. Manag. Care, 2025
[paper]
AppliedPROSPECTClinical prediction models
Optimal vasopressin initiation in septic shock: The OVISS reinforcement learning study
Kalimouttou, Kennedy, Feng, Singh, Saria, Angus, Seymour and Pirracchio
JAMA, 2025
[paper]
AppliedCausal inference
Patterns in symptoms preceding acute care in patients with cancer
Chang, Chen, Feng, Friesner, Mohindra, Boreta, Rabow, Braunstein, Benson and Hong
JAMA Netw. Open, 2025
[paper]
AppliedClinical outcomes research
Who experiences large model decay and why? A Hierarchical Framework for Diagnosing Heterogeneous Performance Drift
Singh, Xia, Gossmann, Chuang, Hong and Feng
International Conference on Machine Learning (ICML), 2025
[paper][code]
MethodsMonitoring & updatingInterpretability
The state of the art in artificial intelligence and digital pathology in prostate cancer
Ni, Kouzy, Sabbagh, Rooney, Feng, Castillo, Gadoue, El Kouzi, Hoffman, Yuan, Madabhushi and Mohamad
Nat. Rev. Urol., 2025
[paper]
Review
Performance drift in a nationally deployed population health risk algorithm in the US Veterans Health Administration
Kolla, Linn, Navathe, Kreisler, Roberts, Park, Singh, Feng, Chen and Parikh
JAMA Health Forum, 2025
[paper]
AppliedMonitoring & updating
Minimally invasive distal pancreatectomy as the standard of care in the US: Are we there yet?
Foroutani, Gonzalez, Wang, Aburayya, Ganjouei, Feng, Thornblade, Hirose, Maker, Nakakura, Corvera, Alseidi and Adam
Cancers (Basel), 2025
[paper]
AppliedClinical outcomes research
Bayesian Concept Bottleneck Models with LLM Priors
Feng, Kothari, Zier, Singh and Tan
Adv. Neural Inf. Process. Syst., 2025
[paper][code][poster]
MethodsInterpretabilityLLMs & generative AI
Potential for algorithmic bias in clinical decision instrument development
Obra, Singh, Watkins, Feng, Obermeyer and Kornblith
NPJ Digit. Med., 2025
[paper]
AppliedEvaluation & auditing
2024
Designing monitoring strategies for deployed machine learning algorithms: navigating performativity through a causal lens
Feng, Subbaswamy, Gossmann, Singh, Sahiner, Kim, Pennello, Petrick, Pirracchio and Xia
Conference on Causal Learning and Reasoning, 2024
[paper][code][poster]
MethodsMonitoring & updatingCausal inference
Monitoring machine learning (ML)-based risk prediction algorithms in the presence of confounding medical interventions
Feng, Gossmann, Pennello, Petrick, Sahiner and Pirracchio
AISTATS, 2024
[paper][code]
MethodsMonitoring & updatingCausal inference
Is this model reliable for everyone? Testing for strong calibration
Feng, Gossmann, Pirracchio, Petrick, Pennello and Sahiner
AISTATS, 2024
[paper][code][poster] Oral presentation
MethodsEvaluation & auditing
Machine Learning Improves Prediction of Postoperative Outcomes After Gastrointestinal Surgery: A Systematic Review and Meta-analysis
Wang, Tozzi, Ganjouei, Romero-Hernandez, Feng, Calthorpe, Castro, Davis, Withers, Zhou, Chaudhary, Adam, Berrevoet, Alseidi and Rashidian
J. Gastrointest. Surg., 2024
[paper]
ReviewClinical prediction models
Machine Learning–Based Prediction of Hospitalization During Chemoradiotherapy With Daily Step Counts
Friesner, Feng, Kalnicki, Garg, Ohri and Hong
JAMA Oncol, 2024
[paper]
AppliedClinical prediction models
A hierarchical decomposition for explaining ML performance discrepancies
Singh, Xia, Subbaswamy, Gossmann and Feng
Adv. Neural Inf. Process. Syst., 2024
[paper][code]
MethodsMonitoring & updatingInterpretability
Transcriptomic profiling of human myocardium at sudden death to define vulnerable substrate for lethal arrhythmias
Caudal, Liu, Pang, Maison, Nakasuka, Feng, Schwarzer-Sperber, Schwarzer, Moffatt, Henrich, Padmanabhan, Connolly, Wu and Tseng
JACC Clin. Electrophysiol., 2024
[paper]
AppliedComputational biology
2023
A Brief Tutorial on Sample Size Calculations for Fairness Audits
Singh, Xia, Kim, Pirracchio, Chunara and Feng
Workshop on Regulatable Machine Learning at the 37th Conference on Neural Information Processing Systems, 2023
[paper]
MethodsEvaluation & auditing
Towards a Post-Market Monitoring Framework for Machine Learning-based Medical Devices: A case study
Feng, Subbaswamy, Gossmann, Singh, Sahiner, Kim, Pennello, Petrick, Pirracchio and Xia
Workshop on Regulatable Machine Learning at the 37th Conference on Neural Information Processing Systems, 2023
[paper][poster][slides] Oral presentation
MethodsAppliedMonitoring & updatingRegulatory science
Development and Validation of Prediction Models and Risk Calculators for Post-Hepatectomy Liver Failure and Postoperative Complications using a Diverse International Cohort of Major Hepatectomies
Wang, Feng, Gomes, Calthorpe, Ashraf Ganjouei, Romero-Hernandez, Benedetti Cacciaguerra, Hibi, Abdelgadir Adam, Alseidi, Abu Hilal, Rashidian and International Post-Hepatectomy Liver Failure Study Group
Ann. Surg., 2023
[paper]
AppliedClinical prediction models
Multi-institutional Development and External Validation of a Machine Learning Model for the Prediction of Distant Metastasis in Patients Treated by Salvage Radiotherapy for Biochemical Failure After Radical Prostatectomy
Sabbagh, Tilki, Feng, Huland, Graefen, Wiegel, B{"o}hmer, Hong, Valdes, Cowan, Cooperberg, Feng, Mohammad, Shelan, D’Amico, Carroll and Mohamad
European Urology Focus, 2023
[paper]
AppliedClinical prediction models
Development of a Machine Learning Model of Postoperative Acute Kidney Injury Using Non-Invasive Time-Sensitive Intraoperative Predictors
Zamirpour, Hubbard, Feng, Butte, Pirracchio and Bishara
Bioengineering, 2023
[paper]
AppliedClinical prediction models
Hand-Assisted Versus Pure Minimally-Invasive Distal Pancreatectomy: Is There a Downside to Lending a Hand?
Miller, Romero-Hernandez, Conroy, Calthorpe, Yilma, Mohamedaly, Kelly, Feng, Hirose, Kirkwood, Maker, Corvera, Nakakura, Alseidi and Adam
World J. Surg., 2023
[paper]
AppliedClinical outcomes research
Development and External Validation of a Machine Learning Model for Prediction of Lymph Node Metastasis in Patients with Prostate Cancer
Sabbagh, Washington, Tilki, Hong, Feng, Valdes, Chen, Wu, Huland, Graefen, Wiegel, Bohmer, Cowan, Cooperberg, Feng, Roach, Trock, Partin, D’Amico, Carroll and Mohamad
Eur Urol Oncol, 2023
[paper]
AppliedClinical prediction models
Hypertension Requiring Medication Use: a Silent Predictor of Poor Outcomes After Pancreaticoduodenectomy
Lin, Conroy, Romero-Hernandez, Yilma, Feng, Hirose, Nakakura, Maker, Corvera, Kirkwood, Alseidi and Adam
J. Gastrointest. Surg., 2023
[paper]
AppliedClinical outcomes research
Predictors of All-Cause Thirty-Day Readmissions in Patients with Heart Failure at an Urban Safety Net Hospital: The Importance of Social Determinants of Health
Steverson, Marano, Chen, Ma, Stern, Feng, Gennatas, Marks, Durstenfeld, Davis, Hsue and Zier
American Journal of Medicine Open, 2023
[paper]
AppliedPROSPECTClinical prediction models
Homelessness and Incidence and Causes of Sudden Death: Data From the POST SCD Study
Haghighat, Ramakrishna, Salazar, Feng, Chiang, Moffatt and Tseng
JAMA Intern. Med., 2023
[paper]
AppliedClinical outcomes research
2022
Bayesian logistic regression for online recalibration and revision of risk prediction models with performance guarantees
Feng, Gossmann, Sahiner and Pirracchio
Journal of the American Medical Informatics Association, 2022
[paper][code]
MethodsMonitoring & updating
Ensembled sparse-input hierarchical networks for high-dimensional datasets
Feng and Simon
Statistical Analysis and Data Mining, 2022
[paper][code]
MethodsHigh-dimensional statistics
Overcoming barriers in the design and implementation of clinical trials for Acute Kidney Injury: a report from the 2020 Kidney Disease Clinical Trialists meeting
Lazzareschi, Mehta, Dember, Bernholz, Turan, Sharma, Kheterpal, Parikh, Ali, Schulman, Ryan, Feng, Simon, Pirracchio, Rossignol and Legrand
Nephrol. Dial. Transplant, 2022
[paper]
PerspectiveClinical outcomes research
Clinical artificial intelligence quality improvement: towards continual monitoring and updating of AI algorithms in healthcare
Feng, Phillips, Malenica, Bishara, Hubbard, Celi and Pirracchio
npj Digital Medicine, 2022
[paper]
PerspectiveMonitoring & updating
Prostate cancer therapy personalization via multi-modal deep learning on randomized phase III clinical trials
Esteva, Feng, van der Wal, Huang, Simko, DeVries, Chen, Schaeffer, Morgan, Sun, Ghorbani, Naik, Nathawani, Socher, Michalski, Roach, Pisansky, Monson, Naz, Wallace, Ferguson, Bahary, Zou, Lungren, Yeung, Ross, Sandler, Tran, Spratt, Pugh, Feng and Mohamad
npj Digital Medicine, 2022
[paper]
AppliedClinical prediction models
2021
Approval policies for modifications to Machine Learning-Based Software as a Medical Device: A study of bio-creep
Feng, Emerson and Simon
Biometrics, 2021
[paper][code]
MethodsMonitoring & updatingRegulatory science
Estimation of cell lineage trees by maximum-likelihood phylogenetics
Feng, DeWitt, McKenna, Simon, Willis and Matsen
Annals of Applied Statistics, 2021
[paper][code]
MethodsComputational biology
2020
Efficient nonparametric statistical inference on population feature importance using Shapley values
Williamson and Feng
International Conference on Machine Learning (ICML), 2020
[paper][code]
MethodsInterpretability
An analysis of the cost of hyper-parameter selection via split-sample validation, with applications to penalized regression
Feng and Simon
Statistica Sinica, 2020
[paper]
MethodsHigh-dimensional statistics