Importance of sample size on the quality and utility of AI-based prediction models for healthcare

Riley, Richard D; Ensor, Joie; Snell, Kym I E; Archer, Lucinda; Whittle, Rebecca; Dhiman, Paula; Alderman, Joseph; Liu, Xiaoxuan; Kirton, Laura; Manson-Whitton, Jay; Smeden, Maarten van; Moons, Karel G.M.; Nirantharakumar, Krishnarajah; Cazier, Jean‐Baptiste; Denniston, Alastair K.; Calster, Ben Van; Collins, Gary S.

doi:10.1016/j.landig.2025.01.013

reviewThe Lancet Digital HealthJun 1, 2025GOLD OA

Importance of sample size on the quality and utility of AI-based prediction models for healthcare

RDRichard D Riley JEJoie Ensor KIKym I E Snell LALucinda Archer RWRebecca Whittle

National Institute for Health Research · NIHR Birmingham Liver Biomedical Research Unit · +10 more institutions

PubMed

Indexed incrossrefdoajpubmed

Abstract

Rigorous study design and analytical standards are required to generate reliable findings in healthcare from artificial intelligence (AI) research. One crucial but often overlooked aspect is the determination of appropriate sample sizes for studies developing AI-based prediction models for individual diagnosis or prognosis. Specifically, the number of participants and outcome events required in datasets for model training and evaluation remains inadequately addressed. Most AI studies do not provide a rationale for their chosen sample sizes and frequently rely on datasets that are inadequate for training or evaluating a clinical prediction model. Among the ten principles of Good Machine Learning Practice…

Citation impact

43

total citations

FWCI: 20.59
Percentile: 100%
References: 64

Citations per year

Authors

17

Topics & keywords

Topics

Keywords

Sample (material)
Sample size determination
Health care
Computer science
Quality (philosophy)
Agency (philosophy)
Artificial intelligence
Outcome (game theory)

No related works found for this paper.