Extended Data Fig. 3: Sequence classification subgroup analysis.
From: A generalizable foundation model for analysis of human brain MRI
Performance comparison of four models, BrainIAC, BrainSegFounder, MedicalNet, and Scratch, on MRI sequence classification tasks across varying training data sizes (10-100% of available data). a One-versus-all classification performance for each MRI sequence type against all other sequences. b Pairwise classification performance between specific sequence pairs. All evaluations use balanced accuracy as the metric.
