Journal of Hepatology. 2023;78(6):119-136 DOI: 10.1016/j.jhep.2023.01.006.

Artificial intelligence, machine learning, and deep learning in liver transplantation

Bhat M, Rabindranath M, Sordi Chara B, Simonetto DA.

Summary
Liver transplantation (LT) is a life-saving treatment for individuals with end-stage liver disease. The management of LT recipients is complex, predominantly because of the need to consider demographic, clinical, laboratory, pathology, imaging, and omics data in the development of an appropriate treatment plan. Current methods to collate clinical information are susceptible to some degree of subjectivity; thus, clinical decision-making in LT could benefit from the data-driven approach offered by artificial intelligence (AI). Machine learning and deep learning could be applied in both the pre- and post-LT settings. Some examples of AI applications pre-transplant include optimising transplant candidacy decision-making and donor-recipient matching to reduce waitlist mortality and improve post-transplant outcomes. In the post-LT setting, AI could help guide the management of LT recipients, particularly by predicting patient and graft survival, along with identifying risk factors for disease recurrence and other associated complications. Although AI shows promise in medicine, there are limitations to its clinical deployment which include dataset imbalances for model training, data privacy issues, and a lack of available research practices to benchmark model performance in the real world. Overall, AI tools have the potential to enhance personalised clinical decision-making, especially in the context of liver transplant medicine.

Keywords

• liver graft
• transplantation
• survival
• waitlist
• mortality

Keypoints

• By integrating large, complex health data, machine and deep learning-based models have the potential to improve outcome predictions compared to conventional biostatistical modelling.
• Organ allocation and donor-recipient matching could be optimised by considering both waitlist mortality and predicted post-transplant outcomes.
• Integrating pre-transplant donor and recipient characteristics could provide a better insight into post-transplant patient and graft survival, informing organ allocation and post-transplant management.
• Management of post-transplant recipients requires identification of risk factors and stratification of patients based on their risk of developing graft-related and other metabolic complications.
• AI predictions may be improved by synthesising various data types which include clinical, imaging, and omics data.
• The main obstacles to AI implementation in the clinic are the need for large, clean, and prospective datasets for model development, model explainability and physician trust, adequate regulatory frameworks for data sharing, reproducible methods, and new research standards that can accurately benchmark performance in the real world.

Introduction
Artificial intelligence (AI) represents a growing area of computer science with widespread acceptance and adoption across numerous disciplines, particularly e-commerce, media, and finance. In the health sciences, the inclusion of machine learning (ML), a field of AI dedicated to performing complex tasks and data analysis, has been slow and is only recently garnering attention.
ML models are trained or learn from datasets through mathematical functions or sets of rules with the delivery of classification and prediction outputs, often with high precision.[1]
Deep learning is a subset of ML based on deep neural networks (DNNs) in which data is analysed through multiple layers of interconnected artificial neurons, simulating the cerebral cortex. ML differs from classical statistics due to its capacity to efficiently model non-linear complex relationships.
The widespread use of electronic health records (EHRs) and storage of immense amounts of longitudinal health data has sparked a growing interest in the development of predictive ML models in medicine, including in the field of solid organ transplantation. This state-of-the-art narrative review will summarise studies that have applied ML techniques in liver transplantation (LT) and provide insight into current limitations and promising directions.

Pre-transplant
Due to the ever-growing disparity between organ supply and demand and the range of factors involved in liver transplant outcomes, new technologies have been explored to improve transplant risk assessment and current allocation systems. ML algorithms can process numerous variables (or features) from large datasets and potentially establish complex relationships between donor and recipient characteristics to complement clinical decision-making in LT (Fig. 1).

Front. Artif. Intell., 15 November 2022 Sec. Medicine and Public Health Volume 5 – 2022 | https://doi.org/10.3389/frai.2022.1050439

Artificial intelligence applied to omics data in liver diseases: Enhancing clinical predictions

Baciu C, Xu C, Alim M, Prayitno K, Bhat M

Abstract

Rapid development of biotechnology has led to the generation of vast amounts of multi-omics data, necessitating the advancement of bioinformatics and artificial intelligence to enable computational modeling to diagnose and predict clinical outcome. Both conventional machine learning and new deep learning algorithms screen existing data unbiasedly to uncover patterns and create models that can be valuable in informing clinical decisions. We summarized published literature on the use of AI models trained on omics datasets, with and without clinical data, to diagnose, risk-stratify, and predict survivability of patients with non-malignant liver diseases. A total of 20 different models were tested in selected studies. Generally, the addition of omics data to regular clinical parameters or individual biomarkers improved the AI model performance. For instance, using NAFLD fibrosis score to distinguish F0-F2 from F3-F4 fibrotic stages, the area under the curve (AUC) was 0.87. When integrating metabolomic data by a GMLVQ model, the AUC drastically improved to 0.99. The use of RF on multi-omics and clinical data in another study to predict progression of NAFLD to NASH resulted in an AUC of 0.84, compared to 0.82 when using clinical data only. A comparison of RF, SVM and kNN models on genomics data to classify immune tolerant phase in chronic hepatitis B resulted in AUC of 0.8793–0.8838 compared to 0.6759–0.7276 when using various serum biomarkers. Overall, the integration of omics was shown to improve prediction performance compared to models built only on clinical parameters, indicating a potential use for personalized medicine in clinical setting.

2024

• Cherry Xu- Best mini-conference presentations- T-CAIREM Summer Student Research presentations – 2024 Summer Student Research Program

2023

• Michael Cooper – GI Research Day 2023 – June 14th, 2023 – Toronto, Canada – (Best Presentation) DYNAMELD: Accurate, Equitable Modeling Of End-Stage Liver Disease

2022

• Fakhar Ali Qazi Arisar – American Society of Transplantation – Living Donor Community of Practice travel award – American Transplant Congress – 2022, June 4 – 8, Boston, MA, USA.
• Fakhar Ali Qazi Arisar – Young Investigator Award – Joint International Congress of ILTS, ELITA & LICAGE – 2022, May 4 – 7, Istanbul, Turkey.
• Nadia Prayitno – EASL NAFLD Summit 2022 Young Investigator Bursary
• Nadia Prayitno – CDTRP Rising Star Travel Award to the Annual Scientific Meeting 2022
• Anh Thu Nguyen-Lefebvre – The Stem Cell Network Travel Award 2022 (TMM conference)
• Michael Cooper – Sheila Sherlock Liver Day 2022 (Best Presentation)  Project Co-supervised with Rahul Krishnan Annual Sheila Sherlock Hepatology Research day.

2021

• Fakhar Ali Qazi Arisar – Early Career Investigator Award in Clinical/Translational Science – American Association for the study of liver diseases (AASLD) – The Liver Meeting Digital Experience (TLMdX) 2021, November 12 – 15.

2020

• Fakhar Ali Qazi Arisar – AASLD Presidential Poster of Distinction – American Association for the study of liver diseases (AASLD) – The Liver Meeting Digital Experience (TLMdX) 2020, November 13 – 16.