Cognitively Biased Prompt Effects on Large Language Model Accuracy for Radiology Board–style Examination Questions

Radiology Artificial Intelligence 2026

View at publisher
Authors: Nicholas Dietrich, Dhruv Patel, Joseph Bellissimo, Christopher T. Loh, Pascal N. Tyrrell
DOI: 10.1148/ryai.250585
Cited by: 1
References: 17

Abstract

Large language model accuracy for radiology board–style questions was reduced by prompts that mimicked cognitive biases, especially on image-based tasks, and prompt-level mitigations, although effective, only partially restored accuracy.

Fields of Study & Subjects

Artificial Intelligence in Healthcare and Education Radiology practices and education Clinical Reasoning and Diagnostic Skills

Publication details

Journal / source
Radiology Artificial Intelligence
Publisher
Radiological Society of North America (RSNA)
Volume
8
Issue
3
Pages
e250585–e250585
Publication date
15 April 2026
Language
EN