Connectome-wide functional connectivity machine learning distinguishes spinal muscular atrophy from healthy controls

Machine learning on connectome-wide functional connectivity discriminated spinal muscular atrophy from controls (43 patients, 45 controls). Key connections in associative cortex and cerebellum. Right parietal–inferior temporal connectivity correlated with motor function.

• Classification performance for SMA vs. healthy controls was stable across logistic regression, support vector machine, and random forest models.
• The most discriminative functional connections involved high-level associative cortical regions and the cerebellum.
• Functional connectivity between right superior parietal gyrus and right inferior temporal gyrus was positively correlated with motor function.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.