A Machine Learning Approach to Liver Histological Evaluation Predicts Clinically Significant Portal Hypertension in NASH Cirrhosis

AI-Cirrhosis-ECG (ACE) score for predicting decompensation and liver outcomes

Background & Aims

Accurate prediction of disease severity and prognosis are challenging in patients with cirrhosis. We evaluated whether the deep learning-based AI-Cirrhosis-ECG (ACE) score could detect hepatic decompensation and predict clinical outcomes in cirrhosis.

Methods

We analyzed 2,166 ECGs from 472 patients in a retrospective Mayo Clinic cohort, 420 patients in a prospective Mayo transplant cohort, and 341 patients in an external validation cohort from Hospital Clínic de Barcelona. The ACE score's performance was assessed using receiver-operating characteristic analysis for decompensation detection and competing risks Cox regression for outcome prediction.

Results

The ACE score showed high accuracy in detecting hepatic decompensation (area under the curve 0.933, 95% CI: 0.923-0.942) with 88.0% sensitivity and 84.3% specificity at an optimal threshold of 0.25. In multivariable analysis, each 0.1-point increase in ACE score was independently associated with increased risk of liver-related death (hazard ratio [HR] 1.44, 95% CI 1.32-1.58, p <0.001). Adding ACE to model for end-stage liver disease-sodium significantly improved prediction of adverse outcomes across all cohorts (c-statistics: retrospective cohort 0.903 vs. 0.844; prospective cohort 0.779 vs. 0.735; external validation 0.744 vs. 0.732; all p <0.001).

Conclusions

The ACE score accurately identifies hepatic decompensation and independently predicts liver-related outcomes in cirrhosis. This non-invasive tool enhances current prognostic models and may improve risk stratification in cirrhosis management.

Impact and implications

This study demonstrates the potential of artificial intelligence to enhance prognostication in liver disease, addressing the critical need for improved risk stratification in cirrhosis management. The AI-Cirrhosis-ECG (ACE) score, derived from widely available ECGs, shows promise as a non-invasive tool for detecting hepatic decompensation and predicting liver-related outcomes, which could significantly impact clinical decision-making and resource allocation in hepatology. These findings are particularly important for hepatologists, transplant surgeons, and patients with cirrhosis, as they offer a novel approach to complement existing prognostic models such as model for end-stage liver disease-sodium. In practical terms, the ACE score could be integrated into routine clinical assessments to provide more accurate risk predictions, potentially improving the timing of interventions, optimizing transplant listing decisions, and ultimately enhancing patient outcomes. However, further validation in diverse populations and integration with other established predictors is necessary before widespread clinical implementation.

Preventing the progression of cirrhosis to decompensation and death

Two main stages are differentiated in patients with advanced chronic liver disease (ACLD), one compensated (cACLD) with an excellent prognosis, and the other decompensated (dACLD), defined by the appearance of complications (ascites, variceal bleeding and hepatic encephalopathy) and associated with high mortality. Preventing the progression to dACLD might dramatically improve prognosis and reduce the burden of care associated with ACLD. Portal hypertension is a major driver of the transition from cACLD to dACLD, and a portal pressure of ≥10 mmHg defines clinically significant portal hypertension (CSPH) as the threshold from which decompensating events may occur. In recent years, innovative studies have provided evidence supporting new strategies to prevent decompensation in cACLD. These studies have yielded major advances, including the development of noninvasive tests (NITs) to identify patients with CSPH with reasonable confidence, the demonstration that aetiological therapies can prevent disease progression and even achieve regression of cirrhosis, and the finding that non-selective β-blockers can effectively prevent decompensation in patients with cACLD and CSPH, mainly by reducing the risk of ascites, the most frequent decompensating event. Here, we review the evidence supporting new strategies to manage cACLD to prevent decompensation and the caveats for their implementation, from patient selection using NITs to ancillary therapies.

Decoding pathology: the role of computational pathology in research and diagnostics

Traditional histopathology, characterized by manual quantifications and assessments, faces challenges such as low-throughput and inter-observer variability that hinder the introduction of precision medicine in pathology diagnostics and research. The advent of digital pathology allowed the introduction of computational pathology, a discipline that leverages computational methods, especially based on deep learning (DL) techniques, to analyze histopathology specimens. A growing body of research shows impressive performances of DL-based models in pathology for a multitude of tasks, such as mutation prediction, large-scale pathomics analyses, or prognosis prediction. New approaches integrate multimodal data sources and increasingly rely on multi-purpose foundation models. This review provides an introductory overview of advancements in computational pathology and discusses their implications for the future of histopathology in research and diagnostics.

Machine learning-based models for advanced fibrosis in non-alcoholic steatohepatitis patients: A cohort study

BACKGROUND

The global prevalence of non-alcoholic steatohepatitis (NASH) and its associated risk of adverse outcomes, particularly in patients with advanced liver fibrosis, underscores the importance of early and accurate diagnosis.

AIM

To develop a machine learning-based diagnostic model for advanced liver fibrosis in NASH patients.

METHODS

A total of 749 patients who underwent liver biopsy at Beijing Ditan Hospital, Capital Medical University, between January 2010 and January 2020 were included. Patients were randomly divided into training (n = 522) and validation (n = 224) cohorts. Five machine learning models were applied to predict advanced liver fibrosis, with feature selection based on Shapley Additive Explanations (SHAP). The diagnostic performance of these models was compared to traditional scores such as the aspartate aminotransferase to platelet ratio index (APRI) and fibrosis index based on the 4 factors (FIB-4), using metrics including the area under the receiver operating characteristic curve (AUROC), decision curve analysis (DCA), and calibration curves.

RESULTS

The Extreme Gradient Boosting (XGBoost) model outperformed all other machine learning models, achieving an AUROC of 0.934 (95%CI: 0.914-0.955) in the training cohort and 0.917 (95%CI: 0.880-0.953) in the validation cohort (P < 0.001). Incorporating liver stiffness measurement into the model further improved its performance, with an AUROC of 0.977 (95%CI: 0.966-0.980) in the training cohort and 0.970 (95%CI: 0.950-0.990) in the validation cohort, significantly surpassing APRI and FIB-4 scores (P < 0.001). The XGBoost model also demonstrated superior clinical utility, as evidenced by DCA and calibration curve analysis in both cohorts.

CONCLUSION

The XGBoost model provides a highly accurate, non-invasive diagnosis of advanced liver fibrosis in NASH patients, outperforming traditional methods. An online tool based on this model has been developed to assist clinicians in evaluating the risk of advanced liver fibrosis.

Development and Validation of a Novel Model to Discriminate Idiosyncratic Drug-Induced Liver Injury and Autoimmune Hepatitis

BACKGROUND AND AIM

Discriminating between idiosyncratic drug-induced liver injury (DILI) and autoimmune hepatitis (AIH) is critical yet challenging. We aim to develop and validate a machine learning (ML)-based model to aid in this differentiation.

METHODS

This multicenter cohort study utilised a development set from Beijing Friendship Hospital, with retrospective and prospective validation sets from 10 tertiary hospitals across various regions of China spanning January 2009 to May 2023. Different ML algorithms were tested using 24 routine laboratory parameters. The Shapley Additive exPlanations (SHAP) analysis was used to evaluate the contribution of each parameter in the ML model.

RESULTS

A total of 2554 patients (1750 for DILI and 804 for AIH) were included. Using Gradient Boost Decision Tree algorithm, five key parameters-aspartate transaminase, globulin, prealbumin, creatinine and platelet count-were selected to construct the ML model. Consequently, a web-based tool named Beijing-AID (BJ-AID) was developed (http://43.143.153.225:5000/). The BJ-AID model demonstrated excellent discrimination performance, with an area under the receiver operating characteristic curve (AUROC) of 0.94 (95% CI, 0.902-0.975) in the development set, 0.91 (95% CI, 0.900-0.928) in all external validation sets and 0.93 (95% CI, 0.889-0.974) in a prospective validation set. Notably, the BJ-AID model also effectively discriminated atypical cases, including drug-induced autoimmune-like hepatitis and AIH with the history of drug consumption, achieving an AUROC = 0.85 (95% CI, 0.742-0.949).

CONCLUSIONS

We successfully developed and validated a machine learning-based model, BJ-AID, which exhibits a strong discrimination performance. BJ-AID can assist practitioners and hepatologists in diagnosing both typical and atypical cases of DILI and AIH.

TRIAL REGISTRATION

ClinicalTrials.gov identifier: NCT05532345.

A machine learning tool for early identification of celiac disease autoimmunity

Identifying which patients should undergo serologic screening for celiac disease (CD) may help diagnose patients who otherwise often experience diagnostic delays or remain undiagnosed. Using anonymized outpatient data from the electronic medical records of Maccabi Healthcare Services, we developed and evaluated five machine learning models to classify patients as at-risk for CD autoimmunity prior to first documented diagnosis or positive serum tissue transglutaminase (tTG-IgA). A train set of highly seropositive (tTG-IgA > 10X ULN) cases (n = 677) with likely CD and controls (n = 176,293) with no evidence of CD autoimmunity was used for model development. Input features included demographic information and commonly available laboratory results. The models were then evaluated for discriminative ability as measured by AUC on a distinct set of highly seropositive cases (n = 153) and controls (n = 41,087). The highest performing model was XGBoost (AUC = 0.86), followed by logistic regression (AUC = 0.85), random forest (AUC = 0.83), multilayer perceptron (AUC = 0.80) and decision tree (AUC = 0.77). Contributing features for the XGBoost model for classifying a patient as at-risk for undiagnosed CD autoimmunity included signs of anemia, transaminitis and decreased high-density lipoprotein. This model's ability to distinguish cases of incident CD autoimmunity from controls shows promise as a potential clinical tool to identify patients with increased risk of having undiagnosed celiac disease in the community, for serologic screening.

AI in Hepatology: Revolutionizing the Diagnosis and Management of Liver Disease

The integration of artificial intelligence (AI) into hepatology is revolutionizing the diagnosis and management of liver diseases amidst a rising global burden of conditions like metabolic-associated steatotic liver disease (MASLD). AI harnesses vast datasets and complex algorithms to enhance clinical decision making and patient outcomes. AI's applications in hepatology span a variety of conditions, including autoimmune hepatitis, primary biliary cholangitis, primary sclerosing cholangitis, MASLD, hepatitis B, and hepatocellular carcinoma. It enables early detection, predicts disease progression, and supports more precise treatment strategies. Despite its transformative potential, challenges remain, including data integration, algorithm transparency, and computational demands. This review examines the current state of AI in hepatology, exploring its applications, limitations, and the opportunities it presents to enhance liver health and care delivery.

Digital quantitation of bridging fibrosis and septa reveals changes in natural history and treatment not seen with conventional histology

BACKGROUND AND AIMS

Metabolic dysfunction-associated steatohepatitis (MASH) with bridging fibrosis is a critical stage in the evolution of fatty liver disease. Second harmonic generation/two-photon excitation fluorescence (SHG/TPEF) microscopy with artificial intelligence (AI) provides sensitive and reproducible quantitation of liver fibrosis. This methodology was applied to gain an in-depth understanding of intra-stage fibrosis changes and septa analyses in a homogenous, well-characterised group with MASH F3 fibrosis.

METHODS

Paired liver biopsies (baseline [BL] and end of treatment [EOT]) of 57 patients (placebo, n = 17 and tropifexor n = 40), with F3 fibrosis stage at BL according to the clinical research network (CRN) scoring, were included. Unstained sections were examined using SHG/TPEF microscopy with AI. Changes in liver fibrosis overall and in five areas of liver lobules were quantitatively assessed by qFibrosis. Progressive, regressive septa, and 12 septa parameters were quantitatively analysed.

RESULTS

qFibrosis demonstrated fibrosis progression or regression in 14/17 (82%) patients receiving placebo, while the CRN scoring categorised 11/17 (65%) as 'no change'. Radar maps with qFibrosis readouts visualised quantitative fibrosis dynamics in different areas of liver lobules even in cases categorised as 'No Change'. Measurement of septa parameters objectively differentiated regressive and progressive septa (p < .001). Quantitative changes in individual septa parameters (BL to EOT) were observed both in the 'no change' and the 'regression' subgroups, as defined by the CRN scoring.

CONCLUSION

SHG/TPEF microscopy with AI provides greater granularity and precision in assessing fibrosis dynamics in patients with bridging fibrosis, thus advancing knowledge development of fibrosis evolution in natural history and in clinical trials.

Image-Guided Liver Biopsy: Perspectives from Interventional Radiology

Liver biopsy is a crucial aspect of interventional radiology and plays a significant role in the management of hepatobiliary diseases. Radiologists commonly perform two major image-guided liver biopsy techniques: percutaneous and transjugular approaches. It is essential for radiologists to understand the role of liver biopsy in diagnosing and treating hepatobiliary conditions, the procedural details involved, and how to manage potential complications. This article reviews the indications, contraindications, techniques, and efficacy of image-guided liver biopsy, with a focus on both percutaneous and transjugular methods.

AI-based automation of enrollment criteria and endpoint assessment in clinical trials in liver diseases

Clinical trials in metabolic dysfunction-associated steatohepatitis (MASH, formerly known as nonalcoholic steatohepatitis) require histologic scoring for assessment of inclusion criteria and endpoints. However, variability in interpretation has impacted clinical trial outcomes. We developed an artificial intelligence-based measurement (AIM) tool for scoring MASH histology (AIM-MASH). AIM-MASH predictions for MASH Clinical Research Network necroinflammation grades and fibrosis stages were reproducible (κ = 1) and aligned with expert pathologist consensus scores (κ = 0.62-0.74). The AIM-MASH versus consensus agreements were comparable to average pathologists for MASH Clinical Research Network scores (82% versus 81%) and fibrosis (97% versus 96%). Continuous scores produced by AIM-MASH for key histological features of MASH correlated with mean pathologist scores and noninvasive biomarkers and strongly predicted progression-free survival in patients with stage 3 (P < 0.0001) and stage 4 (P = 0.03) fibrosis. In a retrospective analysis of the ATLAS trial (NCT03449446), responders receiving study treatment showed a greater continuous change in fibrosis compared with placebo (P = 0.02). Overall, these results suggest that AIM-MASH may assist pathologists in histologic review of MASH clinical trials, reducing inter-rater variability on trial outcomes and offering a more sensitive and reproducible measure of patient responses.

Artificial intelligence scoring of liver biopsies in a phase II trial of semaglutide in nonalcoholic steatohepatitis

BACKGROUND AND AIMS

Artificial intelligence-powered digital pathology offers the potential to quantify histological findings in a reproducible way. This analysis compares the evaluation of histological features of NASH between pathologists and a machine-learning (ML) pathology model.

APPROACH AND RESULTS

This post hoc analysis included data from a subset of patients (n=251) with biopsy-confirmed NASH and fibrosis stage F1-F3 from a 72-week randomized placebo-controlled trial of once-daily subcutaneous semaglutide 0.1, 0.2, or 0.4 mg (NCT02970942). Biopsies at baseline and week 72 were read by 2 pathologists. Digitized biopsy slides were evaluated by PathAI's NASH ML models to quantify changes in fibrosis, steatosis, inflammation, and hepatocyte ballooning using categorical assessments and continuous scores. Pathologist and ML-derived categorical assessments detected a significantly greater percentage of patients achieving the primary endpoint of NASH resolution without worsening of fibrosis with semaglutide 0.4 mg versus placebo (pathologist 58.5% vs. 22.0%, p < 0.0001; ML 36.9% vs. 11.9%; p =0.0015). Both methods detected a higher but nonsignificant percentage of patients on semaglutide 0.4 mg versus placebo achieving the secondary endpoint of liver fibrosis improvement without NASH worsening. ML continuous scores detected significant treatment-induced responses in histological features, including a quantitative reduction in fibrosis with semaglutide 0.4 mg versus placebo ( p =0.0099) that could not be detected using pathologist or ML categorical assessment.

CONCLUSIONS

ML categorical assessments reproduced pathologists' results of histological improvement with semaglutide for steatosis and disease activity. ML-based continuous scores demonstrated an antifibrotic effect not measured by conventional histopathology.

BiliQML: a supervised machine-learning model to quantify biliary forms from digitized whole slide liver histopathological images

The progress of research focused on cholangiocytes and the biliary tree during development and following injury is hindered by limited available quantitative methodologies. Current techniques include two-dimensional standard histological cell-counting approaches, which are rapidly performed, error prone, and lack architectural context or three-dimensional analysis of the biliary tree in opacified livers, which introduce technical issues along with minimal quantitation. The present study aims to fill these quantitative gaps with a supervised machine-learning model (BiliQML) able to quantify biliary forms in the liver of anti-keratin 19 antibody-stained whole slide images. Training utilized 5,019 researcher-labeled biliary forms, which following feature selection, and algorithm optimization, generated an F score of 0.87. Application of BiliQML on seven separate cholangiopathy models [genetic (Afp-CRE;Pkd1l1null/Fl, Alb-CRE;Rbp-jkfl/fl, and Albumin-CRE;ROSANICD), surgical (bile duct ligation), toxicological (3,5-diethoxycarbonyl-1,4-dihydrocollidine), and therapeutic (Cyp2c70-/- with ileal bile acid transporter inhibition)] allowed for a means to validate the capabilities and utility of this platform. The results from BiliQML quantification revealed biological and pathological differences across these seven diverse models, indicating a highly sensitive, robust, and scalable methodology for the quantification of distinct biliary forms. BiliQML is the first comprehensive machine-learning platform for biliary form analysis, adding much-needed morphologic context to standard immunofluorescence-based histology, and provides clinical and basic science researchers with a novel tool for the characterization of cholangiopathies.NEW & NOTEWORTHY BiliQML is the first comprehensive machine-learning platform for biliary form analysis in whole slide histopathological images. This platform provides clinical and basic science researchers with a novel tool for the improved quantification and characterization of biliary tract disorders.

Artificial intelligence-assisted digital pathology for non-alcoholic steatohepatitis: current status and future directions

The worldwide prevalence of non-alcoholic steatohepatitis (NASH) is increasing, causing a significant medical burden, but no approved therapeutics are currently available. NASH drug development requires histological analysis of liver biopsies by expert pathologists for trial enrolment and efficacy assessment, which can be hindered by multiple issues including sample heterogeneity, inter-reader and intra-reader variability, and ordinal scoring systems. Consequently, there is a high unmet need for accurate, reproducible, quantitative, and automated methods to assist pathologists with histological analysis to improve the precision around treatment and efficacy assessment. Digital pathology (DP) workflows in combination with artificial intelligence (AI) have been established in other areas of medicine and are being actively investigated in NASH to assist pathologists in the evaluation and scoring of NASH histology. DP/AI models can be used to automatically detect, localise, quantify, and score histological parameters and have the potential to reduce the impact of scoring variability in NASH clinical trials. This narrative review provides an overview of DP/AI tools in development for NASH, highlights key regulatory considerations, and discusses how these advances may impact the future of NASH clinical management and drug development. This should be a high priority in the NASH field, particularly to improve the development of safe and effective therapeutics.

Liver fibrosis phenotyping and severity scoring by quantitative image analysis of biopsy slides

BACKGROUND & AIMS

Digital pathology image analysis can phenotype liver fibrosis using histological traits that reflect collagen content, morphometry and architecture. Here, we aimed to calculate fibrosis severity scores to quantify these traits.

METHODS

Liver biopsy slides were categorised by Ishak stage and aetiology. We used a digital pathology technique to calculate four fibrosis severity scores: Architecture Composite Score (ACS), Collagen Composite Score (CCS), Morphometric Composite Score (MCS) and Phenotypic Fibrosis Composite Score (PH-FCS). We compared how these scores varied according to disease stage and aetiology.

RESULTS

We included 80 patients (40% female, mean age 59.0 years, mean collagen proportionate area 17.1%) with mild (F0-2, n = 28), moderate (F3-4, n = 17) or severe (F5-6, n = 35) fibrosis. All four aetiology independent scores corelated with collagen proportionate area (ACS: rp  = .512, CCS: rp  = .727, MCS: rp  = .777, PFCS: r = .772, p < .01 for all) with significant differences between moderate and severe fibrosis (p < .05). ACS increased primarily between moderate and severe fibrosis (by 95% to 226% depending on underlying aetiology), whereas MCS and CCS accumulation was more varied. We used 28 qFTs that distinguished between autoimmune- and alcohol-related liver disease to generate an MCS that significantly differed between mild and severe fibrosis for these aetiologies (p < .05).

CONCLUSIONS

We describe four aetiology-dependent and -independent severity scores that quantify fibrosis architecture, collagen content and fibre morphometry. This approach provides additional insight into how progression of architectural changes and accumulation of collagen may differ depending on underlying disease aetiology.

Recent Advances in the Pathogenesis and Clinical Evaluation of Portal Hypertension in Chronic Liver Disease

In chronic liver disease, hepatic stellate cell activation and degeneration of liver sinusoidal endothelial cells lead to structural changes, which are secondary to fibrosis and the presence of regenerative nodules in the sinusoids, and to functional changes, which are related to vasoconstriction. The combination of such changes increases intrahepatic vascular resistance and causes portal hypertension. The subsequent increase in splanchnic and systemic hyperdynamic circulation further increases the portal blood flow, thereby exacerbating portal hypertension. In clinical practice, the hepatic venous pressure gradient is the gold-standard measure of portal hypertension; a value of ≥10 mm Hg is defined as clinically significant portal hypertension, which is severe and is associated with the risk of liver-related events. Hepatic venous pressure gradient measurement is somewhat invasive, so evidence on the utility of risk stratification by elastography and serum biomarkers is needed. The various stages of cirrhosis are associated with different outcomes. In viral hepatitis-related cirrhosis, viral suppression or elimination by nucleos(t)ide analog or direct-acting antivirals results in recompensation of liver function and portal pressure. However, careful follow-up should be continued, because some cases have residual clinically significant portal hypertension even after achieving sustained virologic response. In this study, we reviewed the current and future prospects for portal hypertension.

Predicting the risk of variceal rehemorrhage in cirrhotic patients with portal vein thrombosis: A two-center retrospective study

OBJECTIVES

Although portal vein thrombosis (PVT) was thought to deteriorate portal hypertension and contribute to poor prognosis, risk stratification remains unclear. This study aimed to evaluate its effect on the risk of variceal rehemorrhage and to develop a competitive risk model in cirrhotic patients with PVT.

METHODS

Cirrhotic patients with and without PVT admitted for acute variceal hemorrhage were retrospectively included after matching (1:1) for age, gender and etiology of cirrhosis from two tertiary centers with 1-year follow-up. Those with PVT were subsequently divided into the training and validation cohorts. Cox regression analysis was performed to identify risk factors and develop a competitive risk model, of which the predictive performance and optimal decision threshold were evaluated by C-index, competitive risk curves, calibration curves and decision curve analysis.

RESULTS

Among 398 patients, PVT significantly increased the variceal rehemorrhage risk. Multivariate Cox regression analysis identified that the Child-Turcotte-Pugh score (P = 0.013), chronic PVT (P = 0.025), C-reactive protein (P < 0.001), and aspartate aminotransferase (P = 0.039) were independently associated with variceal rehemorrhage, which were incorporated into the competitive risk model, with high C-index (0.804 and 0.742 of the training and validation cohorts, respectively), risk stratification ability, and consistency. The optimal decision range of the threshold probability was 0.2-1.0.

CONCLUSION

We confirmed the adverse effect of PVT on variceal rehemorrhage and developed a competitive risk model for variceal rehemorrhage in cirrhotic patients with PVT, which might be conveniently used for clinical decision-making.

Deep Neural Network Regression to Assist Non-Invasive Diagnosis of Portal Hypertension

Portal hypertension is a complex medical condition characterized by elevated blood pressure in the portal venous system. The conventional diagnosis of such disease often involves invasive procedures such as liver biopsy, endoscopy, or imaging techniques with contrast agents, which can be uncomfortable for patients and carry inherent risks. This study presents a deep neural network method in support of the non-invasive diagnosis of portal hypertension in patients with chronic liver diseases. The proposed method utilizes readily available clinical data, thus eliminating the need for invasive procedures. A dataset composed of standard laboratory parameters is used to train and validate the deep neural network regressor. The experimental results exhibit reasonable performance in distinguishing patients with portal hypertension from healthy individuals. Such performances may be improved by using larger datasets of high quality. These findings suggest that deep neural networks can serve as useful auxiliary diagnostic tools, aiding healthcare professionals in making timely and accurate decisions for patients suspected of having portal hypertension.

Exploring Opportunities to Enhance the Screening and Surveillance of Hepatocellular Carcinoma in Non-Alcoholic Fatty Liver Disease (NAFLD) through Risk Stratification Algorithms Incorporating Ultrasound Elastography

Non-alcoholic fatty liver disease (NAFLD), with its progressive form, non-alcoholic steatohepatitis (NASH), has emerged as a significant public health concern, affecting over 30% of the global population. Hepatocellular carcinoma (HCC), a complication associated with both cirrhotic and non-cirrhotic NAFLD, has shown a significant increase in incidence. A substantial proportion of NAFLD-related HCC occurs in non-cirrhotic livers, highlighting the need for improved risk stratification and surveillance strategies. This comprehensive review explores the potential role of liver ultrasound elastography as a risk assessment tool for HCC development in NAFLD and highlights the importance of effective screening tools for early, cost-effective detection and improved management of NAFLD-related HCC. The integration of non-invasive tools and algorithms into risk stratification strategies could have the capacity to enhance NAFLD-related HCC screening and surveillance effectiveness. Alongside exploring the potential advancement of non-invasive tools and algorithms for effectively stratifying HCC risk in NAFLD, we offer essential perspectives that could enable readers to improve the personalized assessment of NAFLD-related HCC risk through a more methodical screening approach.

Artificial Intelligence Applications in Hepatology

Over the past 2 decades, the field of hepatology has witnessed major developments in diagnostic tools, prognostic models, and treatment options making it one of the most complex medical subspecialties. Through artificial intelligence (AI) and machine learning, computers are now able to learn from complex and diverse clinical datasets to solve real-world medical problems with performance that surpasses that of physicians in certain areas. AI algorithms are currently being implemented in liver imaging, interpretation of liver histopathology, noninvasive tests, prediction models, and more. In this review, we provide a summary of the state of AI in hepatology and discuss current challenges for large-scale implementation including some ethical aspects. We emphasize to the readers that most AI-based algorithms that are discussed in this review are still considered in early development and their utility and impact on patient outcomes still need to be assessed in future large-scale and inclusive studies. Our vision is that the use of AI in hepatology will enhance physician performance, decrease the burden and time spent on documentation, and reestablish the personalized patient-physician relationship that is of utmost importance for obtaining good outcomes.

Estimation of Physiologic Pressures: Invasive and Non-Invasive Techniques, AI Models, and Future Perspectives

The measurement of physiologic pressure helps diagnose and prevent associated health complications. From typical conventional methods to more complicated modalities, such as the estimation of intracranial pressures, numerous invasive and noninvasive tools that provide us with insight into daily physiology and aid in understanding pathology are within our grasp. Currently, our standards for estimating vital pressures, including continuous BP measurements, pulmonary capillary wedge pressures, and hepatic portal gradients, involve the use of invasive modalities. As an emerging field in medical technology, artificial intelligence (AI) has been incorporated into analyzing and predicting patterns of physiologic pressures. AI has been used to construct models that have clinical applicability both in hospital settings and at-home settings for ease of use for patients. Studies applying AI to each of these compartmental pressures were searched and shortlisted for thorough assessment and review. There are several AI-based innovations in noninvasive blood pressure estimation based on imaging, auscultation, oscillometry and wearable technology employing biosignals. The purpose of this review is to provide an in-depth assessment of the involved physiologies, prevailing methodologies and emerging technologies incorporating AI in clinical practice for each type of compartmental pressure measurement. We also bring to the forefront AI-based noninvasive estimation techniques for physiologic pressure based on microwave systems that have promising potential for clinical practice.

Zonated quantification of immunohistochemistry in normal and steatotic livers

Immunohistochemical stains (IHC) reveal differences between liver lobule zones in health and disease, including nonalcoholic fatty liver disease (NAFLD). However, such differences are difficult to accurately quantify. In NAFLD, the presence of lipid vacuoles from macrovesicular steatosis further hampers interpretation by pathologists. To resolve this, we applied a zonal image analysis method to measure the distribution of hypoxia markers in the liver lobule of steatotic livers.The hypoxia marker pimonidazole was assessed with IHC in the livers of male C57BL/6 J mice on standard diet or choline-deficient L-amino acid-defined high-fat diet mimicking NAFLD. Another hypoxia marker, carbonic anhydrase IX, was evaluated by IHC in human liver tissue. Liver lobules were reconstructed in whole slide images, and staining positivity was quantified in different zones in hundreds of liver lobules. This method was able to quantify the physiological oxygen gradient along hepatic sinusoids in normal livers and panlobular spread of the hypoxia in NAFLD and to overcome the pronounced impact of macrovesicular steatosis on IHC. In a proof-of-concept study with an assessment of the parenchyma between centrilobular veins in human liver biopsies, carbonic anhydrase IX could be quantified correctly as well.The method of zonated quantification of IHC objectively quantifies the difference in zonal distribution of hypoxia markers (used as an example) between normal and NAFLD livers both in whole liver as well as in liver biopsy specimens. It constitutes a tool for liver pathologists to support visual interpretation and estimate the impact of steatosis on IHC results.

The role of artificial intelligence in hepatology research and practice

PURPOSE OF REVIEW

The use of artificial intelligence (AI) in examining large data sets has recently gained considerable attention to evaluate disease epidemiology, management approaches, and disease outcomes. The purpose of this review is to summarize the current role of AI in contemporary hepatology practice.

RECENT FINDINGS

AI was found to be diagnostically valuable in the evaluation of liver fibrosis, detection of cirrhosis, differentiation between compensated and decompensated cirrhosis, evaluation of portal hypertension, detection and differentiation of particular liver masses, preoperative evaluation of hepatocellular carcinoma as well as response to treatment and estimation of graft survival in patients undergoing liver transplantation. AI additionally holds great promise in examination of structured electronic health records data as well as in examination of clinical text (using various natural language processing approaches). Despite its contributions, AI has several limitations, including the quality of existing data, small cohorts with possible sampling bias and the lack of well validated easily reproducible models.

SUMMARY

AI and deep learning models have extensive applicability in assessing liver disease. However, multicenter randomized controlled trials are indispensable to validate their utility.

AI-based histologic scoring enables automated and reproducible assessment of enrollment criteria and endpoints in NASH clinical trials

Clinical trials in nonalcoholic steatohepatitis (NASH) require histologic scoring for assessment of inclusion criteria and endpoints. However, guidelines for scoring key features have led to variability in interpretation, impacting clinical trial outcomes. We developed an artificial intelligence (AI)-based measurement (AIM) tool for scoring NASH histology (AIM-NASH). AIM-NASH predictions for NASH Clinical Research Network (CRN) grades of necroinflammation and stages of fibrosis aligned with expert consensus scores and were reproducible. Continuous scores produced by AIM-NASH for key histological features of NASH correlated with mean pathologist scores and with noninvasive biomarkers and strongly predicted patient outcomes. In a retrospective analysis of the ATLAS trial, previously unmet pathological endpoints were met when scored by the AIM-NASH algorithm alone. Overall, these results suggest that AIM-NASH may assist pathologists in histologic review of NASH clinical trials, reducing inter-rater variability on trial outcomes and offering a more sensitive and reproducible measure of patient therapeutic response.

Machine learning liver histology scores correlate with portal hypertension assessments in nonalcoholic steatohepatitis cirrhosis

BACKGROUND AND AIMS

In cirrhotic nonalcoholic steatohepatitis (NASH) clinical trials, primary efficacy endpoints have been hepatic venous pressure gradient (HVPG), liver histology and clinical liver outcomes. Important histologic features, such as septa thickness, nodules features and fibrosis area have not been included in the histologic assessment and may have important clinical relevance. We assessed these features with a machine learning (ML) model.

METHODS

NASH patients with compensated cirrhosis and HVPG ≥6 mm Hg (n = 143) from the Belapectin phase 2b trial were studied. Liver biopsies, HVPG measurements and upper endoscopies were performed at baseline and at end of treatment (EOT). A second harmonic generation/two-photon excitation fluorescence provided an automated quantitative assessment of septa, nodules and fibrosis (SNOF). We created ML scores and tested their association with HVPG, clinically significant HVPG (≥10 mm Hg) and the presence of varices (SNOF-V).

RESULTS

We derived 448 histologic variables (243 related to septa, 21 related to nodules and 184 related to fibrosis). The SNOF score (≥11.78) reliably distinguished CSPH at baseline and in the validation cohort (baseline + EOT) [AUC = 0.85 and 0.74, respectively]. The SNOF-V score (≥0.57) distinguished the presence of varices at baseline and in the same validation cohort [AUC = 0.86 and 0.73, respectively]. Finally, the SNOF-C score differentiated those who had >20% change in HVPG against those who did not, with an AUROC of 0.89.

CONCLUSION

The ML algorithm accurately predicted HVPG, CSPH, the development of varices and HVPG changes in patients with NASH cirrhosis. The use of ML histology model in NASH cirrhosis trials may improve the assessment of key outcome changes.

Assessment of portal hypertension severity using machine learning models in patients with compensated cirrhosis

BACKGROUND & AIMS

In individuals with compensated advanced chronic liver disease (cACLD), the severity of portal hypertension (PH) determines the risk of decompensation. Invasive measurement of the hepatic venous pressure gradient (HVPG) is the diagnostic gold standard for PH. We evaluated the utility of machine learning models (MLMs) based on standard laboratory parameters to predict the severity of PH in individuals with cACLD.

METHODS

A detailed laboratory workup of individuals with cACLD recruited from the Vienna cohort (NCT03267615) was utilised to predict clinically significant portal hypertension (CSPH, i.e., HVPG ≥10 mmHg) and severe PH (i.e., HVPG ≥16 mmHg). The MLMs were then evaluated in individual external datasets and optimised in the merged cohort.

RESULTS

Among 1,232 participants with cACLD, the prevalence of CSPH/severe PH was similar in the Vienna (n = 163, 67.4%/35.0%) and validation (n = 1,069, 70.3%/34.7%) cohorts. The MLMs were based on 3 (3P: platelet count, bilirubin, international normalised ratio) or 5 (5P: +cholinesterase, +gamma-glutamyl transferase, +activated partial thromboplastin time replacing international normalised ratio) laboratory parameters. The MLMs performed robustly in the Vienna cohort. 5P-MLM had the best AUCs for CSPH (0.813) and severe PH (0.887) and compared favourably to liver stiffness measurement (AUC: 0.808). Their performance in external validation datasets was heterogeneous (AUCs: 0.589-0.887). Training on the merged cohort optimised model performance for CSPH (AUCs for 3P and 5P: 0.775 and 0.789, respectively) and severe PH (0.737 and 0.828, respectively).

CONCLUSIONS

Internally trained MLMs reliably predicted PH severity in the Vienna cACLD cohort but exhibited heterogeneous results on external validation. The proposed 3P/5P online tool can reliably identify individuals with CSPH or severe PH, who are thus at risk of hepatic decompensation.

IMPACT AND IMPLICATIONS

We used machine learning models based on widely available laboratory parameters to develop a non-invasive model to predict the severity of portal hypertension in individuals with compensated cirrhosis, who currently require invasive measurement of hepatic venous pressure gradient. We validated our findings in a large multicentre cohort of individuals with advanced chronic liver disease (cACLD) of any cause. Finally, we provide a readily available online calculator, based on 3 (platelet count, bilirubin, international normalised ratio) or 5 (platelet count, bilirubin, activated partial thromboplastin time, gamma-glutamyltransferase, choline-esterase) widely available laboratory parameters, that clinicians can use to predict the likelihood of their patients with cACLD having clinically significant or severe portal hypertension.

The potential role of machine learning in modelling advanced chronic liver disease

The use of artificial intelligence is rapidly increasing in medicine to support clinical decision making mostly through diagnostic and prediction models. Such models derive from huge databases (big data) including a large variety of health-related individual patient data (input) and the corresponding diagnosis and/or outcome (labels). Various types of algorithms (e.g. neural networks) based on powerful computational ability (machine), allow to detect the relationship between input and labels (learning). More complex algorithms, like recurrent neural network can learn from previous as well as actual input (deep learning) and are used for more complex tasks like imaging analysis and personalized (bespoke) medicine. The prompt availability of big data makes that artificial intelligence can provide rapid answers to questions that would require years of traditional clinical research. It may therefore be a key tool to overcome several major gaps in the model of advanced chronic liver disease, mostly transition from mild to clinically significant portal hypertension, the impact of acute decompensation and the role of further decompensation and treatment efficiency. However, several limitations of artificial intelligence should be overcome before its application in clinical practice. Assessment of the risk of bias, understandability of the black boxes developing the models and models' validation are the most important areas deserving clarification for artificial intelligence to be widely accepted from physicians and patients.

Bibliometric-analysis visualization and review of non-invasive methods for monitoring and managing the portal hypertension

Background

Portal hypertension monitoring is important throughout the natural course of cirrhosis. Hepatic venous pressure gradient (HVPG), regarded as the golden standard, is limited by invasiveness and technical difficulties. Portal hypertension is increasingly being assessed non-invasively, and hematological indices, imaging data, and statistical or computational models are studied to surrogate HVPG. This paper discusses the existing non-invasive methods based on measurement principles and reviews the methodological developments in the last 20 years.

Methods

First, we used VOSviewer to learn the architecture of this field. The publications about the non-invasive assessment of portal hypertension were retrieved from the Web of Science Core Collection (WoSCC). VOSviewer 1.6.17.0 was used to analyze and visualize these publications, including the annual trend, the study hotspots, the significant articles, authors, journals, and organizations in this field. Next, according to the cluster analysis result of the keywords, we further retrieved and classified the related studies to discuss.

Results

A total of 1,088 articles or review articles about our topic were retrieved from WoSCC. From 2000 to 2022, the number of publications is generally growing. “World Journal of Gastroenterology” published the most articles (n = 43), while “Journal of Hepatology” had the highest citations. “Liver fibrosis” published in 2005 was the most influential manuscript. Among the 20,558 cited references of 1,088 retrieved manuscripts, the most cited was a study on liver stiffness measurement from 2007. The highest-yielding country was the United States, followed by China and Italy. “Berzigotti, Annalisa” was the most prolific author and had the most cooperation partners. Four study directions emerged from the keyword clustering: (1) the evaluation based on fibrosis; (2) the evaluation based on hemodynamic factors; (3) the evaluation through elastography; and (4) the evaluation of variceal bleeding.

Conclusion

The non-invasive assessment of portal hypertension is mainly based on two principles: fibrosis and hemodynamics. Liver fibrosis is the major initiator of cirrhotic PH, while hemodynamic factors reflect secondary alteration of splanchnic blood flow. Blood tests, US (including DUS and CEUS), CT, and magnetic resonance imaging (MRI) support the non-invasive assessment of PH by providing both hemodynamic and fibrotic information. Elastography, mainly USE, is the most important method of PH monitoring.

Standardizing the histological assessment of late posttransplantation biopsies from pediatric liver allograft recipients

Excellent short-term survival after pediatric liver transplantation (LT) has shifted attention toward the optimization of long-term outcomes. Despite considerable progress in imaging and other noninvasive modalities, liver biopsies continue to be required to monitor allograft health and to titrate immunosuppression. However, a standardized approach to the detailed assessment of long-term graft histology is currently lacking. The aim of this study was to formulate a list of histopathological features relevant for the assessment of long-surviving liver allograft health and to develop an approach for assessing the presence and severity of these features in a standardized manner. Whole-slide digital images from 31 biopsies obtained ≥4 years after transplantation to determine eligibility for an immunosuppression withdrawal trial were selected to illustrate a range of typical histopathological findings seen in children with clinically stable grafts, including those associated with alloantibodies. Fifty histological features were independently assessed and, where appropriate, scored semiquantitatively by six pathologists to determine inter- and intraobserver reproducibility of the histopathological features using unweighted and weighted kappa statistics; the latter metric enabled distinction between minor and major disagreements in parameter severity scoring. Weighted interobserver kappa statistics showed a high level of agreement for various parameters of inflammation, interface activity, fibrosis, and microvascular injury. Intraobserver agreement for these features was even more substantial. The results of this study will help to standardize the assessment of biopsies from long-surviving liver allografts, aid the recognition of important histological features, and facilitate international comparisons and clinical trials aiming to improve outcomes for children undergoing LT.

Tolerance and acceptance of hepatic venous pressure gradient measurement in cirrhosis (CHESS1904): An international multicenter study

Aim

To determine the tolerance and acceptance of hepatic venous pressure gradient (HVPG) measurements in patients with liver cirrhosis.

Methods

This prospective international multicenter study included 271 patients with cirrhosis who were scheduled to undergo HVPG measurement between October 2019 and June 2020. Data related to the tolerance and acceptance of HVPG measurements were collected using descriptive questionnaires.

Results

HVPG measurements were technically successful in all 271 patients, with 141 (52.0%) undergoing HVPG measurement alone. The complication rate was 0.4%. Postoperative pain was significantly lower than preoperative expected pain (p < 0.001) and intraoperative pain (p < 0.001), and intraoperative pain was also significantly lower than preoperative expected pain (p = 0.036). No, mild, moderate, severe, and intolerable discomfort scores were reported by 36.9%, 44.6%, 11.1%, 6.3%, and 0.4% of these patients, respectively, during HVPG measurement and by 54.6% 32.5%, 11.4%, 1.5%, and 0%, respectively, after HVPG measurement. Of these patients, 39.5% had little understanding and 10% had no understanding of the value of HVPG measurement, with 35.1% and 4.1% regarding HVPG measurements as being of little or no help, respectively. Most patients reported that they would definitely (15.5%), probably (46.9%), or possibly (29.9%) choose to undergo additional HVPG measurements again, and 62.7% regarded the cost of the procedure as acceptable.

Conclusion

HVPG measurement was safe and well‐tolerated in patients with cirrhosis, but patient education and communication are warranted to improve the acceptance of this procedure.

Histological assessment based on liver biopsy: the value and challenges in NASH drug development

Nonalcoholic steatohepatitis (NASH) is increasingly recognized as a serious disease that can lead to cirrhosis, hepatocellular carcinoma (HCC), and death. However, there is no effective drug to thwart the progression of the disease. Development of new drugs for NASH is an urgent clinical need. Liver biopsy plays a key role in the development of new NASH drugs. Histological findings based on liver biopsy are currently used as the main inclusion criteria and the primary therapeutic endpoint in NASH clinical trials. However, there are inherent challenges in the use of liver biopsy in clinical trials, such as evaluation reliability, sampling error, and invasive nature of the procedure. In this article, we review the advantages and value of liver histopathology based on liver biopsy in clinical trials of new NASH drugs. We also discuss the challenges and limitations of liver biopsy and identify future drug development directions.

Hepatic fibrosis 2022: Unmet needs and a blueprint for the future

Steady progress over four decades toward understanding the pathogenesis and clinical consequences of hepatic fibrosis has led to the expectation of effective antifibrotic drugs, yet none has been approved. Thus, an assessment of the field is timely, to clarify priorities and accelerate progress. Here, we highlight the successes to date but, more importantly, identify gaps and unmet needs, both experimentally and clinically. These include the need to better define cell-cell interactions and etiology-specific elements of fibrogenesis and their link to disease-specific drivers of portal hypertension. Success in treating viral hepatitis has revealed the remarkable capacity of the liver to degrade scar in reversing fibrosis, yet we know little of the mechanisms underlying this response. Thus, there is an exigent need to clarify the cellular and molecular mechanisms of fibrosis regression in order for therapeutics to mimic the liver's endogenous capacity. Better refined and more predictive in vitro and animal models will hasten drug development. From a clinical perspective, current diagnostics are improving but not always biologically plausible or sufficiently accurate to supplant biopsy. More urgently, digital pathology methods that leverage machine learning and artificial intelligence must be validated in order to capture more prognostic information from liver biopsies and better quantify the response to therapies. For more refined treatment of NASH, orthogonal approaches that integrate genetic, clinical, and pathological data sets may yield treatments for specific subphenotypes of the disease. Collectively, these and other advances will strengthen and streamline clinical trials and better link histologic responses to clinical outcomes.