ABSTRACT
BACKGROUND: Hepatobiliary resections are challenging due to the complex liver anatomy. Three-dimensional printing (3DP) has gained popularity due to its ability to produce anatomical models based on the characteristics of each patient. METHODS: A multicenter study was conducted on complex hepatobiliary tumours. The endpoint was to validate 3DP model accuracy from original image sources for application in the teaching, patient-communication, and planning of hepatobiliary surgery. RESULTS: Thirty-five patients from eight centers were included. Process testing between 3DP and CT/MRI presented a considerable degree of similarity in vascular calibers (0.22 ± 1.8 mm), and distances between the tumour and vessel (0.31 ± 0.24 mm). The Dice Similarity Coefficient was 0.92, with a variation of 2%. Bland-Altman plots also demonstrated an agreement between 3DP and the surgical specimen with the distance of the resection margin (1.15 ± 1.52 mm). Professionals considered 3DP at a positive rate of 0.89 (95%CI; 0.73-0.95). According to student's distribution a higher success rate was reached with 3DP (median:0.9, IQR: 0.8-1) compared with CT/MRI or 3D digital imaging (P = 0.01). CONCLUSION: 3DP hepatic models present a good correlation compared with CT/MRI and surgical pathology and they are useful for education, understanding, and surgical planning, but does not necessarily affect the surgical outcome.
