No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Use of Artificial Intelligence for Liver Diseases: A Survey from the EASL Congress 2024
Abstract
Artificial intelligence (AI) methods enable humans to analyse large amounts of data, which would otherwise not be feasibly quantifiable. This is especially true for unstructured visual and textual data, which can contain invaluable insights into disease. The hepatology research landscape is complex and has generated large amounts of data to be mined. Many open questions can potentially be addressed with existing data through AI methods. However, the field of AI is sometimes obscured by hype cycles and imprecise terminologies. This can conceal the fact that numerous hepatology research groups already use AI methods in their scientific studies. In this review article, we aim to assess the contemporaneous use of AI methods in hepatology in Europe. To achieve this, we systematically surveyed all scientific contributions presented at the EASL Congress 2024. Out of 1,857 accepted abstracts (1,712 posters and 145 oral presentations), 6 presentations (∼4%) and 69 posters (∼4%) utilised AI methods. Of these, 55 posters were included in this review, while the others were excluded due to missing posters or incomplete methodologies. Finally, we summarise current academic trends in the use of AI methods and outline future directions, providing guidance for scientific stakeholders in the field of hepatology.
... AI is increasingly used in medicine to analyze vast amounts of data, aiding in the diagnosis, prognosis, and prediction of liver diseases such as HCC [52]. AI models, particularly machine learning (ML) and deep learning (DL), often outperform conventional models, such as Cox regression, by handling complex, non-linear, and unstructured data [53]. ...
... AI models, particularly machine learning (ML) and deep learning (DL), often outperform conventional models, such as Cox regression, by handling complex, non-linear, and unstructured data [53]. These models use diverse data sources, including electronic health records, imaging, histopathology, and molecular biomarkers, to enhance HCC risk prediction and early detection [52]. Table 4 ...
Purpose of Review
Accurate hepatocellular carcinoma (HCC) risk prediction in patients with chronic hepatitis B (CHB) is essential for optimizing surveillance and treatment strategies. Various risk models have been developed to identify high-risk individuals requiring intensive monitoring, particularly among both untreated and nucleos(t)ide analog (NUC)-treated patients. This review summarizes existing HCC risk prediction models, including traditional and artificial intelligence (AI)-based approaches, while addressing the evolving role of key predictors such as hepatitis B virus (HBV) DNA levels.
Recent Findings
Conventional models for NUC-treated patients often exclude HBV DNA, assuming its suppression with treatment; however, emerging evidence suggests that baseline HBV DNA may have a long-term, non-linear impact on HCC risk. Despite the growing number of predictive models, their clinical adoption remains limited due to heterogeneity in study populations and the lack of standardized guidelines. AI-driven models have demonstrated potential in enhancing predictive accuracy, yet require further validation for clinical implementation. Additionally, with the advent of functional cure strategies, existing models must be refined to assess HCC risk in post-cure patients.
Summary
A standardized, integrated risk prediction model applicable to both untreated and treated CHB populations is needed to improve risk stratification and guide personalized surveillance and therapeutic strategies. Future research should focus on incorporating key predictors, validating AI-based models, and adapting risk assessment frameworks for evolving treatment paradigms to optimize long-term outcomes in CHB patients.
... AI, trained on lesion-specific imaging features, can offer valuable information not only for disease diagnosis but also to estimate disease progression and predict the prognosis, thereby enhancing disease management. Currently, this area has garnered substantial research attention, with numerous studies focused on the development and validation of AI-driven approaches to improve the reliability of liver disease diagnostics [3]. ...
... AI can be useful in several clinical setting such as 1 ⃝ disease prevention, 2 ⃝ early detection of lesions, and 3 ⃝ the selection of effective treatments (Table 1). For instance, lifestyle improvements such as reducing alcohol consumption can prevent the onset of certain diseases and reduce deaths related to alcohol use. ...
Liver disease can significantly impact life expectancy, making early diagnosis and therapeutic intervention critical challenges in medical care. Imaging diagnostics play a crucial role in diagnosing and managing liver diseases. Recently, the application of artificial intelligence (AI) in medical imaging analysis has become indispensable in healthcare. AI, trained on vast datasets of medical images, has sometimes demonstrated diagnostic accuracy that surpasses that of human experts. AI-assisted imaging diagnostics are expected to contribute significantly to the standardization of diagnostic quality. Furthermore, AI has the potential to identify image features that are imperceptible to humans, thereby playing an essential role in clinical decision-making. This capability enables physicians to make more accurate diagnoses and develop effective treatment strategies, ultimately improving patient outcomes. Additionally, AI is anticipated to become a powerful tool in personalized medicine. By integrating individual patient imaging data with clinical information, AI can propose optimal plans for treatment, making it an essential component in the provision of the most appropriate care for each patient. Current reports highlight the advantages of AI in managing liver diseases. As AI technology continues to evolve, it is expected to advance personalized diagnostics and treatments and contribute to overall improvements in healthcare quality.
... However, analysis of these multimodal datasets is currently a major area of focus in the field [215]. The application of generative artificial intelligence (AI) may present a transformative opportunity to advance our understanding and management of MASLD, such as tabular medical data analysis, multiomics data analysis, the generation of synthetic patient data, enhancing the medical imaging field for MASLD, and advancing personalized medicine approaches for MASLD [216,217]. ...
Metabolic dysfunction-associated steatotic liver disease (MASLD), previously known as nonalcoholic fatty liver disease (NAFLD), is the most prevalent chronic liver disease worldwide, with an estimated global prevalence of approximately 30%; however, effective pharmacotherapies are still limited due to its complex pathogenesis and etiology. Therefore, a more thorough understanding of disease pathogenesis is urgently needed. An increasing number of studies suggest that MASLD and its progressive form, metabolic dysfunction-associated steatohepatitis (MASH), are driven by chronic overnutrition, multiple genetic susceptibility factors, and pathogenic consequences, including hepatocyte damage and liver inflammation. Hepatic inflammation is the key event fueling the conversion from simple steatosis to steatohepatitis and fibrosis. Current therapies for MASH, including the recently approved thyroid hormone receptor-beta agonist resmetirom or the available incretin mimetics, mainly target metabolic injury to the liver but not inflammation directly. In this review, we provide an in-depth discussion of current data related to the immunological mechanisms of MASLD and summarize the effects of current and experimental therapies on immunoregulation in MASLD.
... However, there are limited strategies to conduct in-vitro experiments to mimic post-MBS effects on MASLD. Encouragingly, artificial intelligence and other computational models have emerged in recent years to assist diagnosis, prognostic prediction and therapy development in liver diseases (197). These powerful methods can facilitate evolution in MBS research in the near future. ...
Metabolic dysfunction-associated steatotic liver diseases (MASLD) encompass a spectrum of liver disorders primarily defined by hepatic steatosis and cardiometabolic risk factors. MASLD affects over 30% of the global adult population, representing a substantial risk for public health. MASLD can progress from simple steatosis to steatohepatitis (termed MASH) marked by inflammation and fibrosis, potentially leading to cirrhosis and liver cancer. Obesity has been identified as one of the leading risk factors for MASLD, fueling the concept that efficient anti-obesity therapy will halt MASLD progression. Bariatric surgery (also termed metabolic and bariatric surgery, MBS) has become a well-established intervention for class III obesity and class II obesity with comorbidities, offering significant and sustained weight loss. The common MBS procedures (e.g., Roux-en-Y gastric bypass and sleeve gastrectomy) were indeed found to improve MASLD by ameliorating hepatic steatosis, inflammation, fibrosis, and reducing cancer risk in patients with MASLD/MASH. These effects may go beyond a simple consequence of restricted dietary intake, as illustrated by a growing body of studies showing MBS-induced transcriptional program changes in the liver. Given the mounting evidence from preclinical research, clinical trials and advanced biotechnologies emerging in recent years, refined histological and cellular in-depth characterization in MBS-treated MASLD is anticipated. This review summarizes the current knowledge on the pathogenic and therapeutic effects of MBS on MASLD, with the hope of inspiring further mechanistic and translational studies.
... For example, a multi-omics study proved that an LCD rapidly alters gut microbiota, boosts circulating folate levels, and enhances folate-dependent gene expression in MASLD patients' livers, improving liver fat metabolism and providing multiple metabolic benefits [12]. Therefore, the extensive data generated through liver omics research offers unparalleled opportunities for exploring biological systems at scales previously not achievable, thereby facilitating a deeper understanding of disease mechanisms [17,18]. ...
Metabolic dysfunction-associated steatotic liver disease (MASLD) is the leading cause of chronic liver disease globally. A low-carbohydrate diet (LCD) offers benefits to MASLD patients, albeit its exact mechanism is not fully understood. Using public liver transcriptome data from MASLD patients before/after LCD intervention, we applied differential expression analysis and machine learning to identify key genes. We initially identified 162 differentially expressed genes in the GSE107650 dataset. Secondly, employing two machine learning algorithms, we found that PRAMENP, LEAP2, LOC105379013, and argininosuccinate synthetase 1 (ASS1) are potential hub genes. Additionally, protein–protein interaction and single-cell RNA location analyses suggested that ASS1 was the most crucial hub gene. Then, L1000CDS² analysis of the gene-expression signatures was employed for drug repurposing studies. CGP71683, an appetite suppressant, was predicted to improve MASLD and may mimic the ASS1 expression pattern induced by an LCD. Molecular dynamics confirmed spontaneous, stable CGP71683-ASS1 complex formation. Overall, this work based on analysis of machine learning algorithms, essential gene identification, and drug repurposing studies suggested that ASS1 is an essential gene in MASLD and CGP71683 is a potential drug candidate for treating MASLD by targeting ASS1 and mimicking the beneficial effects of an LCD. However, due to the inherent limitations of a purely computational approach, further experimental investigation is necessary to validate the anticipated results.
Advances in big data analytics, precision medicine, and artificial intelligence are transforming hepatology, offering new insights into disease mechanisms, risk stratification, and therapeutic interventions. In this review, we explore how the integration of genetic studies, multi-omics data, and large-scale population cohorts has reshaped our understanding of liver disease, using steatotic liver disease as a prototype for data-driven discoveries in hepatology. We highlight the role of artificial intelligence in identifying patient subgroups, optimizing treatment strategies, and uncovering novel therapeutic targets. Furthermore, we discuss the importance of collaborative networks, open data initiatives, and implementation science in translating these findings into clinical practice. Although data-driven precision medicine holds great promise, its impact depends on structured approaches that ensure real-world adoption.
Hepatitis B virus (HBV) infection represents a significant global public health threat, often progressing to severe complications such as liver cirrhosis and hepatocellular carcinoma (HCC). Despite the World Health Organization’s 2030 challenge to eliminate HBV, achieving a functional cure remains a major milestone in the fight against the disease. However, even with functional cure, the risk of HCC persists, necessitating ongoing surveillance. Hopefully, advances in technological innovations, artificial intelligence, and the establishment of data platforms have enhanced both functional cure prediction and surveillance strategies. This review examines recent advancements in HBV therapies, prediction models for chronic hepatitis B (CHB) functional cure (such as the GOLDEN model) for evaluating these new approaches, and the current state of HCC surveillance. Furthermore, the potential of novel biomarkers is highlighted in the context of precision medicine to improve CHB functional cure and HCC risk assessment.
The Gut Microbiome-Liver-Brain Axis is a relatively novel construct with promising potential to enhance our understanding of Alcohol Use Disorder (AUD), and its therapeutic approaches. Significant alterations in the gut microbiome occur in AUD even before any other systemic signs or symptoms manifest. Prolonged and inappropriate alcohol consumption, by affecting the gut microbiota and gut mucosa permeability, is thought to contribute to the development of behavioral and cognitive impairments, leading to Alcohol-Related Liver Disorders and potentially progressing into alcoholic cirrhosis, which is often associated with severe cognitive impairment related to neurodegeneration, such as hepatic encephalopathy and alcoholic dementia.
The critical role of the gut microbiota is further supported by the efficacy of FDA-approved treatments for hepatic encephalopathy in alcoholic cirrhosis (i.e., lactulose and rifaximin). To stimulate new research, we hypothesize that interactions between a maladaptive stress response and a constitutional predisposition to neurodegeneration underlie the progression of AUD to conditions of Alcohol-Related Clinical Concerns with severe cognitive impairment, which represent a significant and costly burden to society. Early identification of AUD individuals at risk for developing these conditions could help to prioritize integrated therapeutic interventions targeting different substrates of the Gut Microbiome-Liver-Brain axis. Specifically, addiction medications, microbiome modulators, stress-reducing interventions, and, possibly soon, novel agents that reduce hepatic steatosis/fibrosis will be discussed in the context of digitally supported integrated therapeutic approaches. The explicit goal of this AUD treatment performed on the early stage of the disorder would be to reduce the transition from AUD to those conditions of Alcohol-Related Common Clinical Concerns associated with severe cognitive impairment, a strategy recommended for most neurological neurodegenerative disorders.
Background and aims:
Hepatocellular carcinoma (HCC) is a highly fatal tumor, for which early detection and risk stratification is crucial, yet remains challenging. We aimed to develop an interpretable machine-learning framework for HCC risk stratification based on routinely collected clinical data.
Methods:
We leverage data obtained from over 900,000 individuals and 983 cases of HCC across two large-scale population-based cohorts: the UK Biobank study and the "All Of Us Research Program". For all of these patients, clinical data from timepoints years before diagnosis of HCC was available. We integrate data modalities including demographics, electronic health records, lifestyle, routine blood tests, genomics and metabolomics to offer a unique, multi-modal perspective on HCC risk.
Results:
Our random-forest-based model significantly outperforms all publicly available state-of-the-art risk-scores, with an AUROC of 0.88 both for internal and external test sets. We demonstrate robustness of our model across ethnic subgroups, a major advance over previous models with variable performance by ethnicity. Further, we perform extensive feature-importance analysis, showcasing our approach as an interpretable framework. We provide all model weights and an open-source web calculator to facilitate further validation of our model.
Conclusion:
Our study presents a robust and interpretable machine-learning framework for HCC risk stratification, which offers the potential to improve early detection and could ultimately reduce disease burden through targeted interventions.
Background and aim
Artificial intelligence (AI)-powered chatbots, such as Chat Generative Pretrained Transformer (ChatGPT), have shown promising results in healthcare settings. These tools can help patients obtain real-time responses to queries, ensuring immediate access to relevant information. The study aimed to explore the potential use of ChatGPT-generated medical Arabic responses for patients with metabolic dysfunction–associated steatotic liver disease (MASLD).
Methods
An English patient questionnaire on MASLD was translated to Arabic. The Arabic questions were then entered into ChatGPT 3.5 on November 12, 2023. The responses were evaluated for accuracy, completeness, and comprehensibility by 10 Saudi MASLD experts who were native Arabic speakers. Likert scales were used to evaluate: 1) Accuracy, 2) Completeness, and 3) Comprehensibility. The questions were grouped into 3 domains: (1) Specialist referral, (2) Lifestyle, and (3) Physical activity.
Results
Accuracy mean score was 4.9 ± 0.94 on a 6-point Likert scale corresponding to “Nearly all correct.” Kendall’s coefficient of concordance (KCC) ranged from 0.025 to 0.649, with a mean of 0.28, indicating moderate agreement between all 10 experts. Mean completeness score was 2.4 ± 0.53 on a 3-point Likert scale corresponding to “Comprehensive” (KCC: 0.03–0.553; mean: 0.22). Comprehensibility mean score was 2.74 ± 0.52 on a 3-point Likert scale, which indicates the responses were “Easy to understand” (KCC: 0.00–0.447; mean: 0.25).
Conclusion
MASLD experts found that ChatGPT responses were accurate, complete, and comprehensible. The results support the increasing trend of leveraging the power of AI chatbots to revolutionize the dissemination of information for patients with MASLD. However, many AI-powered chatbots require further enhancement of scientific content to avoid the risks of circulating medical misinformation.
This study aimed to develop nomograms for predicting post-hepatectomy liver failure (PHLF) in patients with hepatocellular carcinoma (HCC), using deep learning analysis of Gadoxetic acid-enhanced hepatobiliary (HBP) MRI.
This retrospective study analyzed patients who underwent gadoxetic acid-enhanced MRI and hepatectomy for HCC between 2016 and 2020 at two referral centers. Using a deep learning algorithm, volumes and signal intensities of whole non-tumor liver, expected remnant liver, and spleen were measured on HBP images. Two multivariable logistic regression models were formulated to predict PHLF, defined and graded by the International Study Group of Liver Surgery: one based on whole non-tumor liver measurements (whole liver model) and the other on expected remnant liver measurements (remnant liver model). The models were presented as nomograms and a web-based calculator. Discrimination performance was evaluated using the area under the receiver operating curve (AUC), with internal validation through 1000-fold bootstrapping.
The study included 1760 patients (1395 male; mean age ± standard deviation, 60 ± 10 years), with 137 (7.8%) developing PHLF. Nomogram predictors included sex, gamma-glutamyl transpeptidase, prothrombin time international normalized ratio, platelets, extent of liver resection, and MRI variables derived from the liver volume, liver-to-spleen signal intensity ratio, and spleen volume. The whole liver and the remnant liver nomograms demonstrated strong predictive performance for PHLF (optimism-corrected AUC of 0.78 and 0.81, respectively) and symptomatic (grades B and C) PHLF (optimism-corrected AUC of 0.81 and 0.84, respectively).
Nomograms based on deep learning analysis of gadoxetic acid-enhanced HBP images accurately stratify the risk of PHLF.
Question Can PHLF be predicted by integrating clinical and MRI-derived volume and functional variables through deep learning analysis of gadoxetic acid-enhanced MRI?
Findings Whole liver and remnant liver nomograms demonstrated strong predictive performance for PHLF with the optimism-corrected area under the curve of 0.78 and 0.81, respectively.
Clinical relevance These nomograms can effectively stratify the risk of PHLF, providing a valuable tool for treatment decisions regarding hepatectomy for HCC.
Most clinical information is encoded as free text, not accessible for quantitative analysis. This study presents an open-source pipeline using the local large language model (LLM) "Llama 2" to extract quantitative information from clinical text and evaluates its performance in identifying features of decompensated liver cirrhosis. The LLM identified five key clinical features in a zero- and one-shot manner from 500 patient medical histories in the MIMIC IV dataset. We compared LLMs of three sizes and various prompt engineering approaches, with predictions compared against ground truth from three blinded medical experts. Our pipeline achieved high accuracy, detecting liver cirrhosis with 100% sensitivity and 96% specificity. High sensitivities and specificities were also yielded for detecting ascites (95%, 95%), confusion (76%, 94%), abdominal pain (84%, 97%), and shortness of breath (87%, 97%) using the 70 billion parameter model, which outperformed smaller versions. Our study successfully demonstrates the capability of locally deployed LLMs to extract clinical information from free text with low hardware requirements.
This paper presents a comprehensive review of the use of Artificial Intelligence (AI) in Systematic Literature Reviews (SLRs). A SLR is a rigorous and organised methodology that assesses and integrates prior research on a given topic. Numerous tools have been developed to assist and partially automate the SLR process. The increasing role of AI in this field shows great potential in providing more effective support for researchers, moving towards the semi-automatic creation of literature reviews. Our study focuses on how AI techniques are applied in the semi-automation of SLRs, specifically in the screening and extraction phases. We examine 21 leading SLR tools using a framework that combines 23 traditional features with 11 AI features. We also analyse 11 recent tools that leverage large language models for searching the literature and assisting academic writing. Finally, the paper discusses current trends in the field, outlines key research challenges, and suggests directions for future research. We highlight three primary research challenges: integrating advanced AI solutions, such as large language models and knowledge graphs, improving usability, and developing a standardised evaluation framework. We also propose best practices to ensure more robust evaluations in terms of performance, usability, and transparency. Overall, this review offers a detailed overview of AI-enhanced SLR tools for researchers and practitioners, providing a foundation for the development of next-generation AI solutions in this field.
