Skull-Face Matching Based on Principal Component Distance of Anatomical Features
JI Da-feng · Biomedical Journal of Scientific & Technical Research · 2024
Objective: To investigate the value of principal component distance of anatomical features in cranial-face matching.Methods: Human head CT data were 50 cases, including 25 males and 25 females.Each skull and facial tissue were reconstructed and stored in the skull and head database, respectively.Six anatomical features in the skull were determined, coordinate points were recorded, and the covariance and principal component coefficients of coordinate points were calculated.The mean and standard deviation of the principal components are calculated by finding the coordinates of the markers, and the mean and the standard deviation are saved in order as a database as the digital mapping of skull-face.The Euclidean distance is compared between the skull principal component coefficient and the principal component coefficient of the database skull model to be matched, and the face model mapped by the minimum distance is taken as the result of the skull face simulation.Rebel the 40 skull anatomical feature points in the model library and match the skull model in the database to verify the reliability and robustness of the data algorithm.Results: The correct number of matches in the database was 37 (92.5%). Conclusion:The principal component distance based on anatomical features can effectively match the facial data corresponding to the skull.