Domain-specific functional outcome prediction in stroke rehabilitation: A multicenter artificial intelligence study.

Our findings demonstrate the predictive feasibility of artificial intelligence-based modeling for domain-specific stroke prognosis and present a prototype implementation illustrating the potential clinical applicability of the proposed framework across heterogeneous institutional settings.

The use of routinely collected clinical variables and the preliminary cross-institutional validation results support further investigation of real-world rehabilitation practices, pending prospective clinical evaluation.