Abstract

Modern hepatobiliary surgery relies on an increasingly precise understanding of hepatic anatomy. Three- Dimensional (3D) anatomical reconstruction from medical imaging has enabled the transition from static, 2D anatomy to dynamic computational anatomy. This article reviews how advanced processing of CT and MRI images, enhanced by Deep Learning, is improving morphological studies, making them more individualized. It analyzes how 3D hepatic image reconstruction can reveal vascular and biliary variants, as well as complex venous drainage maps that are difficult to assess in 2D images. The results of recent studies addressing anatomical variants, the classification of portal branching, venous congestion in oncological resections, the accuracy of liver volume calculation, and improved surgical outcomes are discussed. “Virtual 3D anatomy” should currently be the standard for obtaining the most precise hepatic anatomy in the living patient.

Keywords:3D reconstruction; Computational anatomy; Anatomical variants; Hepatic segmentation; Volumetry; Venous maps; CT

Keywords: 3D/2D: Three-Dimensional/Two-Dimensional; CT: Computed Tomography (Spanish: TC); RLV: Remnant Liver Volume (Spanish: VHR); CNN: Convolutional Neural Networks; LDLT: Living Donor Liver Transplantation; HCC: Hepatocellular Carcinoma; AR: Augmented Reality (Spanish: RA)