“Despite the clear benefits of using AI to analyse complex data types in hepatology, there are still substantial barriers and challenges to its integration into routine clinical workflows. In this position paper, we assessed the limitations and propose clear recommendations to ease the transition of AI-based diagnostic, prognostic or predictive systems to improve clinical care.
Professor Debbie Shawcross
18 July 2025
Expert recommendations for the use of AI in the diagnosis, treatment and research of liver diseases
Researchers from King’s College London are part of a group of 34 international leading experts in hepatology, data science, and artificial intelligence (AI) that has released a consensus report presenting concrete recommendations for the clinical application of AI in hepatology.
The paper was published today in the Journal of Hepatology by the European Association for the Study of the Liver (EASL) and its AI Task Force, which was established to guide the responsible and effective use of AI in hepatology.
AI systems have rapidly evolved over the past 15 years, with major advances particularly in image-based applications in radiology and pathology – and more recently, also in language-based methods. Despite this progress, clinical use of AI remains limited due to numerous implementation challenges. To better understand these barriers and support the integration of AI tools into hepatology, the EASL AI Task Force conducted a structured consensus process involving expert hepatologists and data scientists, as well as clinical AI and regulatory sciences experts.
The process was led by Professor Jakob N. Kather of Else Kröner Fresenius Center (EKFZ) for Digital Health, TU Dresden and University Hospital Dresden, Germany, Dr Sabela Lens of University Hospital Barcelona, Spain, and Dr. Eric Trépo of University Hospital Brussels, Belgium. They based it on the Delphi method – a systematic, iterative, anonymised survey procedure widely used in medicine to develop guidelines. It ensures that resulting recommendations reflect broad expert consensus.
“Over the past decade, Artificial Intelligence (AI) in the field of hepatology has multiplied at fast pace. Yet, adoption of AI methods in real-world clinical practice and clinical research remains limited,” says Professor Debbie Shawcross, Professor of Hepatology and Chronic Liver Failure in the Roger Williams Institute of Liver Studies at King’s and co-author of the paper.
AI literacy as a key to implementation
A central recommendation of the group is to build AI literacy among healthcare professionals. The expert panel further advocates that AI systems must demonstrate efficacy, reliability, and trustworthiness before they can be adopted into clinical practice. Even validated tools often face practical barriers due to heterogenous infrastructure across healthcare systems and limited interoperability of hospital IT systems. Addressing these implementation challenges is crucial to enabling routine use.
Conducting AI-supported clinical trials and facilitating data exchange
The experts also highlight the importance of investigating AI methods in clinical trials. This requires seamless information exchange between institutions, decentralised access to de-identified data, and early collaboration between AI researchers and clinical trial investigators.
Promoting interdisciplinary collaboration and creating structured frameworks
The paper highlights how professional societies could help, for example by establishing targeted initiatives and structured frameworks to support AI implementation. The experts recommend that future clinical guidelines, coordinated by these societies, should explicitly discuss recommendations of AI models. Those guidelines should further serve as prioritised clinical context for AI-based decision support systems.
The new recommendations aim to assist healthcare institutions, industry stakeholders, and policymakers in advancing the responsible, patient-centred use of AI in hepatology worldwide.
“AI is opening new frontiers in how we understand and treat liver diseases,” adds Dr Ellis Paintsil, co-author and expert in antimicrobial resistance at King’s College London. “It enables us to unravel complex, multi-layered datasets – such as microbiome and resistance profiles – in ways that were previously not possible."
Our new recommendations offer a much-needed blueprint to help transition AI tools from the research environment into clinical settings, especially in complex, data-rich fields like hepatology.
Dr Ellis Paintsil
“What is unique about this publication is the international, interdisciplinary consensus – the result of close collaboration between experts from different countries and disciplines. This common basis can help us to translate pioneering AI technologies from the lab into clinical practice. We illustrate this roadmap for hepatology, but our goal was to develop a scalable blueprint also for other areas of medicine as they face similar challenges,” says Dr Jan Clusmann, postdoctoral researcher at the EKFZ for Digital Health at TU Dresden and first author of the publication.
Read the full position paper here.
In this story
Professor of Hepatology and Chronic Liver Failure