Utilizing Machine Learning to Predict Liver Allograft Fibrosis by Leveraging Clinical and Imaging Data
Clinical Transplantation · 2025
DOI: 10.1111/ctr.70148 ↗Kurz gefasst
Fachartikel in der Zeitschrift Clinical Transplantation, 2025. Verfasst von Madhumitha Rabindranath und 3 weiteren Autor:innen. Beteiligte Einrichtung: Ajmera Transplant Program, University Health Network, Toronto, Ontario, Canada. Hauptthemen laut MEDLINE: Liver Transplantation, Liver Cirrhosis, Machine Learning, Graft Rejection. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Mitacs Accelerate, Canada Graduate Scholarship - Master programs.
Abstract (Originalsprache)
Background and Aim Liver transplant (LT) recipients may succumb to graft‐related pathologies, contributing to graft fibrosis (GF). Current methods to diagnose GF are limited, ranging from procedural‐related complications to low accuracy. With recent advances in machine learning (ML), we aimed to develop a noninvasive tool using demographic, clinical, laboratory, and B‐mode ultrasound (US) features to predict significant fibrosis (METAVIR≥F2).
Methods
We used a nested 10‐fold cross‐validation approach with grid‐search for hyperparameter fine‐tuning to train an artificial neural network (ANN) and a support vector machine (SVM) to classify mild fibrosis (F0‐F1) and significant fibrosis (F2‐F4) on 1131 patients. We calculated Shapley values to identify top‐ranked features, determining the contribution of each feature to model predictions. For the imaging‐based model, we used 4819 images with 892 studies trained on the residual network 18 (ResNet18) model to classify F0‐F1 versus F3‐F4.
Results
We determined the ANN performed the best when compared to the SVM and standard biomarkers, with an AUC ranging from 0.77 to 0.81. The ResNet18 model was unable to diagnose advanced GF, leading to the training AUCs ranging from 0.89 to 0.97, while the validation and testing AUCs were 0.43–0.63. Shapley analysis highlighted the following top‐ranked features associated with significant GF: hepatitis C at transplant, recipient age, recipient sex, and certain blood markers such as creatinine and hemoglobin.
Conclusion
Noninvasive approaches using ML for predicting significant GF perform well when considering demographic, clinical, and laboratory data; however, this performance is not enhanced with the use of US images.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Ajmera Transplant Program, University Health Network, Toronto, Ontario, Canada.
- Institute of Medical Science, University of Toronto, Toronto, Ontario, Canada.
- Toronto General Hospital Research Institute, University Health Network, Toronto, Ontario, Canada.
- Department of Medical Imaging, Princess Margaret Hospital, Toronto, Ontario, Canada.
- Division of Gastroenterology & Hepatology, Department of Medicine, University of Toronto, Toronto, Ontario, Canada.
ORCID-Profile: Madhumitha Rabindranath, Mamatha Bhat
Förderung
- Mitacs Accelerate
- Canada Graduate Scholarship - Master programs
Zitieren
APA
Rabindranath, M., Sun, Y., Khalili, K., & Bhat, M. (2025). Utilizing Machine Learning to Predict Liver Allograft Fibrosis by Leveraging Clinical and Imaging Data. Clinical Transplantation. https://doi.org/10.1111/ctr.70148
Vancouver
Rabindranath M, Sun Y, Khalili K, Bhat M. Utilizing Machine Learning to Predict Liver Allograft Fibrosis by Leveraging Clinical and Imaging Data. Clinical Transplantation. 2025. doi:10.1111/ctr.70148
BibTeX
@article{Rabindranath2025,
author = {Rabindranath, Madhumitha and Sun, Yingji and Khalili, Korosh and Bhat, Mamatha},
title = {Utilizing Machine Learning to Predict Liver Allograft Fibrosis by Leveraging Clinical and Imaging Data},
journal = {Clinical Transplantation},
year = {2025},
doi = {10.1111/ctr.70148}
}
EUVIMED