Minimizing and quantifying uncertainty in AI-informed decisions: Applications in medicine

S Samuel D. Curtis (Department of Pharmacology and Molecular Sciences, Johns Hopkins University School of Medicine) S Sambit Panda (Department of Biomedical Engineering, Johns Hopkins University) A Adam Li (Department of Computer Science, Columbia University) H Haoyin Xu (Department of Biomedical Engineering, Johns Hopkins University) Y Yuxin Bai (Department of Biomedical Engineering, Johns Hopkins University) I Itsuki Ogihara (Department of Biomedical Engineering, Johns Hopkins University) E Eliza O’Reilly (Department of Applied Mathematics and Statistic, Johns Hopkins University) Y Yuxuan Wang L Lisa Dobbyn (Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine) M Maria Popoli (Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine) J Janine Ptak (Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine) N Nadine Nehme (Department of Oncology, the Sidney Kimmel Cancer Center, Johns Hopkins University School of Medicine) N Natalie Silliman (Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine) J Jeanne Tie (Division of Personalized Oncology, Walter and Eliza Hall Institute of Medical Research) P Peter Gibbs (Division of Personalized Oncology, Walter and Eliza Hall Institute of Medical Research) L Lan T. Ho-Pham (BioMedical Research Center, Pham Ngoc Thach University of Medicine) B Bich N. H. Tran (Saigon Precision Medicine Research Center) T Thach S. Tran (Saigon Precision Medicine Research Center) T Tuan V. Nguyen (Saigon Precision Medicine Research Center) E Ehsan Irajizad (Department of Biostatistics, The University of Texas MD Anderson Cancer Center) M Michael Goggins (Department of Oncology, the Sidney Kimmel Cancer Center, Johns Hopkins University School of Medicine) C Christopher L. Wolfgang (Department of Surgery, New York University Langone) T Tian-Li Wang (Department of Pathology, Johns Hopkins University School of Medicine) I Ie-Ming Shih (Department of Pathology, Johns Hopkins University School of Medicine) A Amanda Fader (Department of Gynecology and Obstetrics, Johns Hopkins Medical Institutions) A Anne Marie Lennon (Department of Medicine, University of Pittsburgh Medical Center) R Ralph H. Hruban (Department of Oncology, the Sidney Kimmel Cancer Center, Johns Hopkins University School of Medicine) C Chetan Bettegowda L Lucy Gilbert (Department of Oncology, McGill University Health Centre, Montreal) K Kenneth W. Kinzler (Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine) N Nickolas Papadopoulos (Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine) B Bert Vogelstein (Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine) J Joshua T. Vogelstein (Department of Biomedical Engineering, Johns Hopkins University) C Christopher Douville (Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine)

Abstract

AI is now a cornerstone of modern dataset analysis. In many real world applications, practitioners are concerned with controlling specific kinds of errors, rather than minimizing the overall number of errors. For example, biomedical screening assays may primarily be concerned with mitigating the number of false positives rather than false negatives. Quantifying uncertainty in AI-based predictions, and in particular those controlling specific kinds of errors, remains theoretically and practically challenging. We develop a strategy called multidimensional informed generalized hypothesis testing (MIGHT) which we prove accurately quantifies uncertainty and confidence given sufficient data, and concomitantly controls for particular error types. Our key insight was that it is possible to integrate canonical cross-validation and parametric calibration procedures within a nonparametric ensemble method. Simulations demonstrate that while typical AI based-approaches cannot be trusted to obtain the truth, MIGHT can be. We apply MIGHT to answer an open question in liquid biopsies using circulating cell-free DNA (ccfDNA) in individuals with or without cancer: Which biomarkers, or combinations thereof, can we trust? Performance estimates produced by MIGHT on ccfDNA data have coefficients of variation that are often orders of magnitude lower than other state of the art algorithms such as support vector machines, random forests, and Transformers, while often also achieving higher sensitivity. We find that combinations of variable sets often decrease rather than increase sensitivity over the optimal single variable set because some variable sets add more noise than signal. This work demonstrates the importance of quantifying uncertainty and confidence—with theoretical guarantees—for the interpretation of real-world data.

Article Details

Volume / Issue Vol. 122, Issue 34
Published August 26, 2025
ISSN 0027-8424
Publisher National Academy of Sciences

Authors (34)

S

Samuel D. Curtis

Department of Pharmacology and Molecular Sciences, Johns Hopkins University School of Medicine

S

Sambit Panda

Department of Biomedical Engineering, Johns Hopkins University

A

Adam Li

Department of Computer Science, Columbia University

H

Haoyin Xu

Department of Biomedical Engineering, Johns Hopkins University

Y

Yuxin Bai

Department of Biomedical Engineering, Johns Hopkins University

I

Itsuki Ogihara

Department of Biomedical Engineering, Johns Hopkins University

E

Eliza O’Reilly

Department of Applied Mathematics and Statistic, Johns Hopkins University

Y

Yuxuan Wang

L

Lisa Dobbyn

Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine

M

Maria Popoli

Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine

J

Janine Ptak

Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine

N

Nadine Nehme

Department of Oncology, the Sidney Kimmel Cancer Center, Johns Hopkins University School of Medicine

N

Natalie Silliman

Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine

J

Jeanne Tie

Division of Personalized Oncology, Walter and Eliza Hall Institute of Medical Research

P

Peter Gibbs

Division of Personalized Oncology, Walter and Eliza Hall Institute of Medical Research

L

Lan T. Ho-Pham

BioMedical Research Center, Pham Ngoc Thach University of Medicine

B

Bich N. H. Tran

Saigon Precision Medicine Research Center

T

Thach S. Tran

Saigon Precision Medicine Research Center

T

Tuan V. Nguyen

Saigon Precision Medicine Research Center

E

Ehsan Irajizad

Department of Biostatistics, The University of Texas MD Anderson Cancer Center

M

Michael Goggins

Department of Oncology, the Sidney Kimmel Cancer Center, Johns Hopkins University School of Medicine

C

Christopher L. Wolfgang

Department of Surgery, New York University Langone

T

Tian-Li Wang

Department of Pathology, Johns Hopkins University School of Medicine

I

Ie-Ming Shih

Department of Pathology, Johns Hopkins University School of Medicine

A

Amanda Fader

Department of Gynecology and Obstetrics, Johns Hopkins Medical Institutions

A

Anne Marie Lennon

Department of Medicine, University of Pittsburgh Medical Center

R

Ralph H. Hruban

Department of Oncology, the Sidney Kimmel Cancer Center, Johns Hopkins University School of Medicine

C

Chetan Bettegowda

L

Lucy Gilbert

Department of Oncology, McGill University Health Centre, Montreal

K

Kenneth W. Kinzler

Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine

N

Nickolas Papadopoulos

Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine

B

Bert Vogelstein

Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine

J

Joshua T. Vogelstein

Department of Biomedical Engineering, Johns Hopkins University

C

Christopher Douville

Department of Oncology, the Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine