Multivariate Pattern Analysis Reveals Neurobiological Markers of Communication Impairments in Children with Bilateral Spastic Cerebral Palsy

Motivation: Communication impairments in children with bilateral spastic cerebral palsy (BSCP) are linked to specific brain morphological and connectivity changes, which require individualized diagnostic approaches for effective assessment. Goal(s): To develop a personalized diagnostic model that identifies neurobiological markers of communication impairments in BSCP, enabling targeted interventions. Approach: A multivariate pattern analysis integrating cortical morphology and white matter connectivity metrics was applied to compare BSCP children with typically developing controls. Results: The combined support vector classification (SVC) model, based on cortical and white matter features, achieved 80.7% accuracy and an AUC of 0.881, effectively distinguishing communication impairments in BSCP. Impact: The MVPA-based model facilitates individualized diagnostics for communication impairments in BSCP, supporting personalized treatment and rehabilitation, and paving the way for precision neuroimaging applications in pediatric neurodevelopmental disorders.