Muscular dystrophies are traditionally monitored using peripheral measures of muscle strength and function. However, many patients experience fatigue, cognitive difficulties, sleep disturbance, and reduced daily functioning. These brain-related symptoms contribute substantially to disease burden but remain difficult to assess objectively in both clinical care and clinical trials. We present a clinically motivated, AI-based framework to quantify brain involvement in muscular dystrophy by transforming complex brain data into interpretable biomarkers of central nervous system (CNS) burden. Multimodal data were collected from pediatric and adult patients with muscular dystrophy evaluated in neuromuscular clinics at Stanford University and the University of Minnesota, including diffusion MRI, resting-state functional MRI, electroencephalography (EEG), and standardized cognitive assessments. AI models identify reproducible patterns of brain network disruption associated with patient-reported fatigue, cognitive performance, and functional outcomes. To further probe CNS mechanisms, we performed a targeted multimodal analysis integrating EEG network features with RNA mis-splicing profiles in myotonic dystrophy type 1. EEG analyses revealed disease-specific spectral slowing, altered functional connectivity, and distinct neurophysiological subtypes, while transcriptomic variance was dominated by spliceosomal dysregulation and synaptic pathways. Cross-modal alignment demonstrated convergent disease-related axes linking molecular pathology to systems-level brain dysfunction, supporting EEG as a scalable, non-invasive surrogate for CNS burden and longitudinal monitoring. Model development and validation are further supported by well-characterized resources, enabling robust analysis despite the challenges of rare disease research. To ensure clinical relevance, imaging-derived brain signatures are translated into EEG-based markers that can be acquired repeatedly in outpatient settings, allowing longitudinal monitoring of brain function without reliance on advanced imaging at every visit. This strategy aligns CNS assessment with real-world neuromuscular clinic workflows and supports incorporation of brain-based endpoints into therapeutic trials. AI-based brain biomarkers offer a practical pathway to better capture symptoms that matter most to patients, improve disease monitoring, and support more comprehensive, patient-centered care.