Objective: to develop and validate a weighted heuristic evaluation model designed to measure the usability of augmentative and alternative communication (AAC) systems for individuals with neuromuscular diseases, integrating both expert and end-user perspectives. Methodology: through applied research based on User-Centered Design (UCD), a two-round Delphi method was implemented to validate a final set of 20 usability heuristics categorized into efficiency, effectiveness, learnability, and satisfaction. The instrument was administered to a panel of two human-computer interaction experts and a group of 20 users with neuromuscular disorders (spinal muscular atrophy, amyotrophic lateral sclerosis, myotonic dystrophy, and Duchenne muscular dystrophy). Importance scores provided by users were processed using the median to derive variable weights incorporated into a normalized Usability Index (UI) formula. The global analysis was validated using parametric Student's t-tests and ANOVA. Result: at the macro level, statistical tests showed no significant overall differences between experts and users (p = 0.9091; F = 0.01; p = 0.9226). However, item-by-item granular analysis identified critical priority discrepancies (≥ 0.3) in the dimensions of error correction, few-step execution, adaptability, and, most notably, assistive technology compatibility (0.5 variation). Conclusions: the proposed weighted model provides a rigorous, objective, and replicable mathematical framework that overcomes the biases of traditional heuristic evaluations with uniform weights, demonstrating that technical interoperability prioritized by experts must align with the daily practical experience of users with motor limitations.