Additional file 1 of Concurrent AI-human interaction in prostate cancer MRI interpretation: More hype than help?

Abstract

Additional file 1: Table S1 Technical parameters of the local acquisition protocol. Table S2. Summary of intra- and inter-reader intraclass correlation coefficient (ICC) for patient-level likelihood of clinically significant cancer, reported for overall readings and stratified by radiologist experience. Table S3 Summary of the Generalized Linear Mixed Model (GLMM) to estimate the effect of expertise and AI-reading on the overall likelihood score of csPCa. Table S4 Patient-level detection performance for any prostate cancer, reported overall and stratified by radiologist experience. Table S5. Proportion of Gleason ≥ 7 cancer detected, biopsies performed, Gleason < 7 cancer detected, and Gleason ≥ 7 cancer missed, reported for overall readings and stratified by radiologist experience level across four biopsy recommendation strategies. Fig. S1. Percentage change in the patient-level likelihood of clinically significant cancer according to the experience. Fig. S2 Receiver operating characteristic analysis with diagnostic performance metrics for any prostate cancer, with and without AI assistance, stratified by reader expertise: (a) expert radiologists; (b) basic radiologists; (c) residents; and (d) overall.

Ähnliche Arbeiten