Preoperative prediction of post hepatectomy liver failure after surgery for hepatocellular carcinoma on CT-scan by machine learning and radiomics analyses.
Authors
Affiliations (17)
Affiliations (17)
- Hepatobiliary Surgery Unit, Fondazione Policlinico Universitario A. Gemelli, IRCCS, Catholic University of the Sacred Heart, Rome, Italy; IRCAD, Research Institute Against Cancer of the Digestive System, 1 Place de l'Hôpital, Strasbourg, 67091, France. Electronic address: [email protected].
- Department of Radiology, San Gerardo Hospital, Monza, Italy.
- Department of Hepatobiliary and General Surgery, IRCCS Humanitas Research Hospital, Rozzano, Milan, Italy.
- Hepatobiliary Surgery Unit, Fondazione Policlinico Universitario A. Gemelli, IRCCS, Catholic University of the Sacred Heart, Rome, Italy.
- Division of Hepato-biliary-pancreatic, Minimally Invasive and Robotic Surgery, and Transplantation Service Federico II University HospitalNaplesItaly, Italy.
- School of Medicine and Surgery, University of Milan-Bicocca, Department of Surgery, San Gerardo Hospital, Monza, Italy.
- Division of General and Hepatobiliary Surgery, Department of Surgical Sciences, Dentistry, Gynecology and Pediatrics, University of Verona, Verona, Italy.
- Unit of General Surgery 1, University of Pavia and Foundation IRCCS Policlinico San Matteo, Pavia, Italy.
- General and Oncologic Surgery, Morgagni-Pierantoni Hospital, Department of Medical and Surgical Sciences, University of Bologna Forlì, Italy.
- Department of Surgical, Oncological and Gastroenterological Science (DISCOG), University of Padua, Hepatobiliary and Pancreatic Surgery Unit, Treviso Hospital, Italy.
- Department of General Surgery, Ospedale Carlo Poma, Mantua, Italy.
- Department of General and Pediatric Surgery, Bolzano Central Hospital, Bolzano, Italy.
- Pole d'Expertise de la Regulation Numérique (PEReN), Paris, France.
- IRCAD, Research Institute Against Cancer of the Digestive System, 1 Place de l'Hôpital, Strasbourg, 67091, France.
- Dpt of Surgery, University Hospital of Geneva, Switzerland; ICube lab, Photonics for Health, University of Strasbourg, France.
- HepatoBiliaryPancreatic Surgery, AOU Careggi, Department of Experimental and Clinical Medicine, University of Florence, Florence, Italy.
- Department of Health Sciences, Università del Piemonte Orientale, Novara, Italy; Department of General Surgery, University Maggiore Hospital Della Carità, Novara, Italy.
Abstract
No instruments are available to predict preoperatively the risk of posthepatectomy liver failure (PHLF) in HCC patients. The aim was to predict the occurrence of PHLF preoperatively by radiomics and clinical data through machine-learning algorithms. Clinical data and 3-phases CT scans were retrospectively collected among 13 Italian centres between 2008 and 2022. Radiomics features were extracted in the non-tumoral liver area. Data were split between training(70 %) and test(30 %) sets. An oversampling was run(ADASYN) in the training set. Random-Forest(RF), extreme gradient boosting (XGB) and support vector machine (SVM) models were fitted to predict PHLF. Final evaluation of the metrics was run in the test set. The best models were included in an averaging ensemble model (AEM). Five-hundred consecutive preoperative CT scans were collected with the relative clinical data. Of them, 17 (3.4 %) experienced a PHLF. Two-hundred sixteen radiomics features per patient were extracted. PCA selected 19 dimensions explaining >75 % of the variance. Associated clinical variables were: size, macrovascular invasion, cirrhosis, major resection and MELD score. Data were split in training cohort (70 %, n = 351) and a test cohort (30 %, n = 149). The RF model obtained an AUC = 89.1 %(Spec. = 70.1 %, Sens. = 100 %, accuracy = 71.1 %, PPV = 10.4 %, NPV = 100 %). The XGB model showed an AUC = 89.4 %(Spec. = 100 %, Sens. = 20.0 %, Accuracy = 97.3 %, PPV = 20 %, NPV = 97.3 %). The AEM combined the XGB and RF model, obtaining an AUC = 90.1 %(Spec. = 89.5 %, Sens. = 80.0 %, accuracy = 89.2 %, PPV = 21.0 %, NPV = 99.2 %). The AEM obtained the best results in terms of discrimination and true positive identification. This could lead to better define patients fit or unfit for liver resection.