2/5 How does rLNM work?
For sensitivity, we move beyond arbitrary thresholds and use an anatomy-preserving spatial null model to control false positives/negatives. For specificity, we regress out cross-symptom effects to reveal cleaner, symptom-specific networks.