A different skilled artist, working with the same skull and the same consultation, might have made somewhat different choices — and might have produced a somewhat different face.

Method Three: AFA3D and the Computational Approach

The third reconstruction was the work of Renitta Rajan Thottungal, working with a computational method called AFA3D — Automated Facial Approximation 3D — developed as part of French research infrastructure including the TIVMI (Traitement d’Images et Vision par ordinateur pour la Médecine et l’Industrie) research group.

AFA3D is a fundamentally different kind of approach from the two manual methods.

Rather than a human expert interpreting skull morphology through their trained judgment, AFA3D is an algorithm that takes a skull scan as input and produces a facial approximation through an automated computational process, by reference to a large statistical database of skull-face relationships derived from CT scans of living individuals.