Artificial intelligence enhances the accuracy of portal and hepatic vein extraction in computed tomography for virtual hepatectomy

Abstract

Background/Purpose Current conventional algorithms used for 3‐dimensional simulation in virtual hepatectomy still have difficulties distinguishing the portal vein (PV) and hepatic vein (HV). The accuracy of these algorithms was compared with a new deep‐learning based algorithm (DLA) using artificial intelligence. Methods A total of 110 living liver donor candidates until 2017, and 46 donor candidates until 2019 were allocated to the training group and validation groups for the DLA, respectively. All PV or HV branches were labeled based on Couinaud's segment classification and the Brisbane 2000 Terminology by hepato‐biliary surgeons. Misclassified and missing branches were compared between a conventional tracking‐based algorithm (TA) and DLA in the validation group. Results The sensitivity, specificity, and Dice coefficient for the PV were 0.58, 0.98, and 0.69 using the TA; and 0.84, 0.97, and 0.90 using the DLA (P < .001, excluding specificity); and for the HV, 0.81, 087, and 0.83 using the TA; and 0.93, 0.94 and 0.94 using the DLA (P < .001 to P = .001). The DLA exhibited greater accuracy than the TA. Conclusion Compared with the TA, artificial intelligence enhanced the accuracy of extraction of the PV and HVs in computed tomography.

Two‐step artificial intelligence algorithm for liver segmentation automates anatomic virtual hepatectomy

“…In the training data, the outer edge of the whole liver, PV, and hepatic vein (HV) branches were extracted from the CT data of each of the 174 donors to develop the first and the second step of the AI algorithm. The first step of the AI algorithm to extract the PV and HV from the CT data has been previously reported 11 (Synapse Vincent, version 6, Fujifilm Corp.). To develop the second step of the AI, 2 experts, hepato‐biliary surgeons with 10 and 9 years of experience (YK and RT, respectively), made corrections and labeled each of the PV and HV branches to define the two hemilivers, four sectors, and eight segments for liver segmentation, as described later.…”

Artificial intelligence‐based technology to make a three‐dimensional pelvic model for preoperative simulation of rectal cancer surgery using MRI

“…Validation of the vascular and nerve segmentation was carried out according to the previous method used in the assessment of portal and hepatic vein extraction for hepatectomy. 12 Namely, the intrapelvic vessels and nerves were labeled by a colorectal surgeon (AH), and whether these anatomies could be segmented successfully was considered. For cases with variant anatomical type, such as lacking an obturator artery branching from the internal iliac artery, the vessels were not labeled.…”

“…These cases were independent of the cases for which ground truth labels were created. Validation of the vascular and nerve segmentation was carried out according to the previous method used in the assessment of portal and hepatic vein extraction for hepatectomy 12 . Namely, the intrapelvic vessels and nerves were labeled by a colorectal surgeon (AH), and whether these anatomies could be segmented successfully was considered.…”

“…Taking these characteristics of 3D MRI into consideration, it is plausible that MRI would be efficient at creating 3D images for preoperative simulation. Until now, no tools have been available to reconstruct 3D images from MRI for the treatment of rectal cancer, though we considered that recent advancements in artificial intelligence (AI) can realize the automated segmentation differently from the classic methods 12,13 . In the present study, we attempted to develop an automated algorithm to create a 3D model from pelvic MRI using AI technology, thereby contributing to advancements in preoperative simulation.…”

Imaging-based deep learning in liver diseases

“…Careful surgery planning with accurate segmentation of major vessels is of great importance before hepatectomy and liver transplantation. Kazami et al [ 9 ] developed a DL algorithm for fast portal vein and hepatic vein segmentation on CT images. The sensitivity and DSC of the DL algorithm were significantly higher than traditional tracking-based algorithm.…”