Browse the corpus
Walk the evidence base by book and chapter — the raw source passages that ground Ask, Differential, and the rest.
1 passage
Artificial intelligence in prediction of non-alcoholic fatty liver disease and fibrosis. Artificial intelligence (AI) has become increasingly widespread in our daily lives, including healthcare applications. AI has brought many new insights into better ways we care for our patients with chronic liver disease, including non-alcoholic fatty liver disease and liver fibrosis. There are multiple ways to apply the AI technology on top of the conventional invasive (liver biopsy) and noninvasive (transient elastography, serum biomarkers, or clinical prediction models) approaches. In this review article, we discuss the principles of applying AI on electronic health records, liver biopsy, and liver images. A few common AI approaches include logistic regression, decision tree, random forest, and XGBoost for data at a single time stamp, recurrent neural networks for sequential data, and deep neural networks for histology and images.