Introduction: Chronic liver diseases are a common source of morbidity and mortality in both children and adults in the United States and around the world. According to the U.S. Centers for Disease Control, chronic liver diseases and cirrhosis had an age-adjusted death rate of 10.8 per 100,000 population. This death rate was higher in the male population (14.5 per 100,000 population), and is the 10th most common overall cause of death. Liver stiffening is most often the result of tissue fibrosis in the setting of chronic liver diseases. As such, MRI elasticity imaging have been developed for detecting and assessing the severity of liver stiffness in children and adults to follow liver disease over time and demonstrate either response to medical therapy or disease progression.
Non-invasive imaging tests, including ultrasound, magnetic resonance imaging (MRI), and computed tomography (CT), have played increasing roles on diagnosis of chronic liver diseases over the past few decades. Elasticity imaging can be performed using either commercially-available ultrasound or MRI equipment and allows quantitative evaluation of liver stiffness. MR elastography, in particular, uses an active-passive driver system (with the passive paddle placed over the right upper quadrant of the abdomen at the level of the costal margin) to create transverse (shear) waves in the liver. The displacement of liver tissue related to these waves can be imaged using a modified phase-contrast pulse sequence and can be used to create an elastogram (map or parametric image) of liver stiffness. Although MR elastography obviates the need for liver biopsy for some patients and allows more frequent longitudinal assessment of liver health, it has associated drawbacks related to additional patient time in the scanner, patient discomfort, and added costs (e.g., infrastructure and patient charge-related).
We developed a multi-channel deep transfer learning model, DeepLiverNet, to categorically classify the severity of liver stiffness using both anatomic T2-weighted MRI and clinical data for pediatric and adult patients with known or suspected pediatric chronic liver diseases. [will be online soon] This deep learning model achieved an accuracy of 88.1% in a cross-validation experiment, and outperformed our previously developed Support Vector Machine classifier. (He, Li et al. 2019) Such a validated model is able to triage the need for additional MREMR elastography testing, and thus potentially avoid MMR elastography in up to two-thirds of candidate patients, shortening examination length, and lowering healthcare costs.
He, L., H. Li, J. A. Dudley, T. C. Maloney, S. L. Brady, E. Somasundaram, A. T. Trout and J. R. Dillman (2019). Machine Learning Prediction of Liver Stiffness Using Clinical and T2-Weighted MRI Radiomic Data. American Journal of Roentgenology 213(3): 1-10. Pubmed Journal
Hailong Li, Lili He, Jonathan Dudley, Thomas Maloney, Elanchezhian Somasundaram, Samuel L. Brady, Nehal A. Parikh, and Jonathan R. Dillman (2020). DeepLiverNet: A Deep Transfer Learning Model for Classifying Liver Stiffness using Clinical and T2-Weighted MRI Data. Pediatric Radiology, In Press.