A Parallel Framework for HCC Detection in DCE-MRI Sequences with Wavelet-Based Description and SVM classification

Abstract

In this article, we propose a complete framework devoted to detect liver HCC (Hepato-Cellular Carcinoma) tumors within DCE-MRI (Dynamic Contrast Enhanced-MRI) sequences. Our system employs different phases of these hepatic image sequences (depending on time after contrast agent injection) to describe local patches with wavelet-based descriptors. By using a SVM (Support Vector Machine)-based classification, we are able to distinguish healthy patches from pathological ones. Moreover, thanks to a parallel image processing strategy, we are able to reduce significantly the running time so that our system may be utilized as a computer aided diagnosis tool in the future. Our experiments show that our contribution is an accurate system for HCC detection, with a small cohort of patients, but representing a high volume of image data to be processed. This work encourages us to conduct deeper researches for detecting complex HCC cases for larger patients cohorts.