One-way analysis of variance (ANOVA) revealed highly significant differences in classification performance among brain regions for both datasets (RepOD: F = 38.48, p < 0.0 .001 ; Kaggle: F = 236.83, p < 0.0 .001 ), providing strong evidence that regional selection significantly impacts ScZ detection accuracy.
← all excerpts
Exploring brain lobe-specific insights in an explainable framework for EEG-based schizophrenia detection.
1
—
—