Program Information

Discrimination of the Liver Cirrhosis and Normal Liver in T1- Weighted MR Images Using Radiomic Features

J Duan

Shandong Cancer Hospital Affiliated to Shandong University, Jinan, Shandong province

Presentations

SU-I-GPD-J-95 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall

Purpose: In this study, radiomic features was proposed to assess liver cirrhosis and normal liver in T1- weighted MR image, and to automatically identify cirrhosis tissue and normal tissue for liver.

Methods: T1- weighted MR images from 15 liver cirrhosis patients in our institution between 2013 and 2016 were randomly selected for this study. Meanwhile, 15 normal liver T1- weighted MR data sets were selected as a control. We manually delineated the cirrhosis areas and the normal areas. All feature calculations were performed using the Imaging Biomarker Explorer (IBEX) software (free open source). To individuate which parameter was available in classifying the cirrhosis and normal liver, a Wilcoxon signed-rank test between textural indices was performed.

Results: In total, we extracted 89 radiomic features that were grouped into 15 feature classes. About 20 parameters (such as correlation, variance, energy, etc) show significant differences in identifying the cirrhosis liver and normal liver.

Conclusion: The results indicate that radiomic features for liver parenchyma maybe helpful in identifying cirrhosis tissue and normal tissue. However, a deeper validation of this new approach on a large sample is required in the future.


Contact Email: