Abstract
<jats:p>This paper presents an artificial intelligence-assisted methodological framework for the design and fabrication of patient-specific cranial plug prostheses from computed tomography data. The proposed workflow integrates AI-assisted semantic segmentation, three dimensional reconstruction, topological repair of the anatomical model, geometric parameter extraction for prosthesis generation, and an expert-in-the-loop validation stage to ensure clinical viability prior to fabrication. The methodology was validated through a comparative analysis against a traditional manual fabrication procedure, using a small-scale cranial perforation as a test case with high geometric fit requirements. Results showed a substantial reduction in fabrication time and an improvement in the anatomical fit of the AI-generated device. Overall, the findings support the proposed framework as a replicable, precise, and efficient alternative for patient-specific prosthesis fabrication in clinically demanding settings.</jats:p>