AI applications for quantitative histopathological analysis of NASH tissue sections

The prevalence of NASH is increasing while no drug treatment is currently available. Liver biopsies remain the gold standard for its diagnosis. Currently, NASH histologic assessment is semi-quantitative using different scores (e.g., SAF, Kleiner…) where measures are scaled (e.g.: from 0 to 3) bringing subjectivity to the analysis. Tribun Health has developed AI applications for quantitative histopathological analysis of NASH tissue sections to standardize and improve the drug development journey. These deep learning-based applications can identify and quantify the relevant histopathological features (steatosis, fibrosis, ballooning, and inflammation) with detailed reports.