Primary sclerosing cholangitis (PSC) is a rare liver disease characterized by progressive multi-focal bile-duct strictures and can ultimately lead to liver cirrhosis. Multimodal MRI and MRCP are usually used to detect PSC-compatible bile duct changes. Since reading MRCP can be challenging and time-consuming, we aimed to develop a deep learning-based decision support system for the automated classification of PSC on 2D-MRCP images. For this, we designed a deep learning ensemble algorithm which leverages and processes the multiple rotational views of the 2D-MRCP in parallel, that even outperforms the average radiologists in predictive accuracy.
Stay In Touch