A Machine Learning Approach Enables Quantitative Measurement of Liver Histology and Disease Monitoring in NASH

Abstract

BaCKgRoUND aND aIMS: Manual histological assessment is currently the accepted standard for diagnosing and monitoring disease progression in NASH, but is limited by variability in interpretation and insensitivity to change. Thus, there is a critical need for improved tools to assess liver pathology in order to risk stratify NASH patients and monitor treatment response. appRoaCH aND ReSUltS: Here, we describe a machine learning (ML)-based approach to liver histology assessment, which accurately characterizes disease severity and heterogeneity, and sensitively quantifies treatment response in NASH. We use samples from three randomized controlled trials to build and then validate deep convolutional neural networks to measure key histological features in NASH, including steatosis, inflammation, hepatocellular ballooning, and fibrosis. The ML-based predictions showed strong correlations with expert pathologists and were prognostic of progression to cirrhosis and liver-related clinical events. We developed a heterogeneity-sensitive metric of fibrosis response, the Deep Learning Treatment Assessment Liver Fibrosis score, which measured antifibrotic treatment effects that went undetected by manual pathological staging and was concordant with histological disease progression. CoNClUSIoNS:Our ML method has shown reproducibility and sensitivity and was prognostic for disease progression, demonstrating the power of ML to advance our understanding of disease heterogeneity in NASH, risk stratify affected patients, and facilitate the development of therapies. (Hepatology 2021;74:133-147). G lobal prevalence of NAFLD is rising rapidly. (1,2) NAFLD represents a continuum of disease of varying severity, with milder forms consisting of simple steatosis, whereas the progressive form, NASH, can progress to cirrhosis and end-stage liver disease. NASH-related cirrhosis is now the fastest growing indication for liver transplantation (LT) in the USA. (3)

Novel artificial intelligent/neural network system for staging of nonalcoholic steatohepatitis

“…Application of ML had been explored for pattern recognition in NAFLD using liver biopsy images, 28 ultrasonography, 29,30 and clinical data 31–33 ; however, these are unsatisfactory for the differential diagnosis between NAFLD and NASH and for staging liver fibrosis in NASH 34 …”

Outcomes and potential surrogate markers for future clinical trials of non‐alcoholic steatohepatitis cirrhosis

“…A recent paper evaluating a machine-learning (ML) based-approach with paired biopsies from 3 large NASH RCTs including patients with advanced fibrosis (STELLAR-3, STELLAR-4 and ATLAS) 60 showed that these artificial intelligence-driven techniques are sensitive and reliable, and represent promising approaches to correlate dynamic changes in fibrosis (even in the F4 stage) with clinical outcomes, although the number of events in those clinical trials was very small and further studies are required to increase the confidence in this novel and exciting techniques. cluding hepatic collagen and fat content and α-SMA expression.…”

“…In an attempt to mitigate the inter and intra‐observer variability in NASH histology readings discussed in the first section, improved machine‐driven methods to quantify basal amount of fibrosis and changes in time are being increasingly explored in the last few years. A recent paper evaluating a machine‐learning (ML) based‐approach with paired biopsies from 3 large NASH RCTs including patients with advanced fibrosis (STELLAR‐3, STELLAR‐4 and ATLAS) 60 showed that these artificial intelligence‐driven techniques are sensitive and reliable, and represent promising approaches to correlate dynamic changes in fibrosis (even in the F4 stage) with clinical outcomes, although the number of events in those clinical trials was very small and further studies are required to increase the confidence in this novel and exciting techniques.…”

Bridge Category Models: Development of Bayesian Modelling Procedures to Account for Bridge Ordinal Ratings for Disease Staging

“…Algorithms in arti cial intelligence (AI) may resolve issues with interrater variability [54], measurement uncertainty and error by providing quantitative assessments of tissue histology and other tangential tasks which may be of immediate value [55,12,56,57,58,59]. A combination of both rater uncertainty and incomplete information about the histology may present additional challenges for training and evaluating AI technologies.…”