Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Review
. 2024 Feb;80(2):335-351.
doi: 10.1016/j.jhep.2023.10.015. Epub 2023 Oct 24.

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

Vlad Ratziu  1 , Marcus Hompesch  2 , Mathieu Petitjean  3 , Cindy Serdjebi  4 , Janani S Iyer  5 , Anil V Parwani  6 , Dean Tai  7 , Elisabetta Bugianesi  8 , Kenneth Cusi  9 , Scott L Friedman  10 , Eric Lawitz  11 , Manuel Romero-Gómez  12 , Detlef Schuppan  13 , Rohit Loomba  14 , Valérie Paradis  15 , Cynthia Behling  16 , Arun J Sanyal  17
Affiliations
Review

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

Vlad Ratziu et al. J Hepatol. 2024 Feb.

Abstract

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.

Keywords: NAFLD; NASH; artificial intelligence; ballooning; clinical trials; digital pathology; fibrosis; histology; inflammation; liver biopsy; machine learning; steatosis.

PubMed Disclaimer

Conflict of interest statement

Conflict of interest

Mathieu Petitjean is an employee of PharmaNest, Inc., Princeton, NJ, USA; Cindy Serdjebi is an employee of Biocellvia, Marseille, France; Janani S. Iyer is an employee of PathAI, Inc., Boston, MA, USA; Dean Tai is an employee of HistoIndex Pte Ltd, Singapore. VR: received consulting fees from Novo-Nordisk, Sagimet, Madrigal, Enyo, Poxel, Northsea, Intercept Pharmaceuticals, Prosciento and research grants (to institution) from Gilead Sciences and Intercept Pharmaceuticals. SLF: Active Consulting: Alnylam, Axcella Health, Cargene, Cellarity, ChemomAb, Fate Therapeutics, Galmed, Gordian Biotechnology, Glycotest, Hepgene, In sitro, Korro Bio, Laekna, Laronde, Ochre Bio, Merck, Morphic Therapeutics, Overtone Therapeutics, North Sea Therapeutics, Pfizer Pharmaceuticals, Pliant, Prosciento Research, Resolution Therapeutics, Sagimet, Satellite Bio, Surrozen, Takeda Pharmaceuticals, Yaqrit; Stock options: Escient, Galectin Galmed, Genfit, Glympse, Gordian Biotechnology, Hepgene, Laekna, Lifemax, Metacrine, Morphic Therapeutics, Nimbus, North Sea, Therapeutics, Sagimet, Satellite Bio, Scholar Rock, Surrozen; Research Collaborations/Research with Commercial Entities: Morphic Therapeutics; Novo Nordisk; Abalone Bio (SBIR Grant); Espervita, Galmed, Pionyr. Please refer to the accompanying ICMJE disclosure forms for further details.

Figures

Fig. 1.
Fig. 1.. AIM-NASH and NASH explore workflow and outputs.
Liver biopsy samples are stained with H&E and MT prior to conventional scanning and digitisation performed by either the user or at PathAI Diagnostics. WSIs are uploaded via the cloud and are evaluated first by artifact detection and exclusion algorithms, and subsequently by AIM-NASH and/or NASH-explore machine learning-based algorithms for tissue characterisation. WSI, heatmap overlays, and quantitative human interpretable features are either returned directly to the user or are reviewed by a pathologist who accepts or rejects the model-derived ordinal scores prior to delivery. Examples of AIM-NASH output images illustrating key NASH histological features are shown in the lower panel and were provided by Dr Janini Iyer of PathAI, Inc. AIM-NASH, AI-based histologic measurement non-alcoholic steatohepatitis; CRN, Clinical Research Network; MT, Masson’s trichrome; NAS, non-Alcoholic fatty liver disease activity score; NASH, non-alcoholic steatohepatitis; WSI, whole-slide imaging.
Fig. 2.
Fig. 2.. FibroNest™ workflow and outputs.
H&E, Masson’s trichrome, PSR, and IHC staining are performed on liver biopsy samples prior to scanning and digitisation. WSIs are uploaded via the cloud and analysed by FibroNest™ machine learning-based algorithms. Images and quantitative data are returned for pathologist review and interpretation. Examples of FibroNest™ output images illustrating key NASH histological features are shown in the lower panel and provided by Dr Mathieu Petitjean of PharmaNest Inc. IHC, immunohistochemistry; NASH, non-alcoholic steatohepatitis; PSR, picrosirius red; WSI, whole-slide images.
Fig. 3.
Fig. 3.. MorphoQuant® workflow and outputs.
H&E, PSR, and IHC staining are performed on liver biopsy samples prior to scanning and digitisation. WSI are uploaded via the cloud and analysed by MorphoQuant® automated algorithms. Images and quantitative data are returned for pathologist review and interpretation. Examples of MorphoQuant® output images illustrating key NASH histological features are shown in the lower panel and provided by Dr Cindy Serdjebi of Biocellvia. IHC, immunohistochemistry; NASH, non-alcoholic steatohepatitis; PSR, picrosirius red; WSI, whole-slide imaging.
Fig. 4.
Fig. 4.. qFibrosis, qInflammation, qBallooning, and qSteatosis workflow and outputs.
Unstained liver biopsy samples are imaged by SHG/TPE fluorescence. WSI are uploaded via the cloud and analysed by qFibrosis knowledge-based algorithms. Images and quantitative data are returned for pathologist review and interpretation. Examples of qFibrosis output images illustrating key NASH histological features are shown in the lower panel and provided by Dr Dean Tai of HistoIndex Pte Inc., as further described in Liu et al. and Naoumov et al., AI, artificial intelligence; AUROC, area under the receiver operating characteristic; MT, Masson’s trichrome; NASH, non-alcoholic steatohepatitis; SHG/TPE, second harmonic generation/two-photon excitation; WSI, whole-slide imaging.
Fig. 5.
Fig. 5.. A proposal for clinical validation of DP/AI diagnostic procedures for NASH.
AI, artificial intelligence; DP, digital pathology; NASH, nonalcoholic steatohepatitis.
Fig. 6.
Fig. 6.. Diagnostic capabilities of DP/AI methodologies applied to liver histology.
AI, artificial intelligence; DP, digital pathology.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Harrison SA, Gawrieh S, Roberts K, et al. Prospective evaluation of the prevalence of non-alcoholic fatty liver disease and steatohepatitis in a large middle-aged US cohort. J Hepatol 2021;75:284–291. - PubMed
    1. Le MH, Le DM, Baez TC, et al. Global incidence of non-alcoholic fatty liver disease: a systematic review and meta-analysis of 63 studies and 1,201,807 persons. J Hepatol 2023;79:287–295. - PubMed
    1. Golabi P, Paik JM, AlQahtani S, et al. Burden of non-alcoholic fatty liver disease in asia, the Middle East and North africa: data from global burden of disease 2009–2019. J Hepatol 2021;75:795–809. - PubMed
    1. Food Usa and Administration Drug. Noncirrhotic nonalcoholic steatohepatitis with liver fibrosis: developing drugs for treatment guidance for industry. Draft Guidance; 2018.
    1. European Medicines Agency. Reflection paper on regulatory requirements for the development of medicinal products for chronic non 6 infectious liver diseases. PBC, PSC, NASH); 2018.

Publication types