A Fully-Automated Pipeline for Detection and Segmentation of Liver Lesions and Pathological Lymph Nodes

TL;DR

This paper introduces a fully-automated pipeline combining machine learning and active contour models for accurate detection and segmentation of liver lesions and lymph nodes in CT images, demonstrating promising preliminary results.

Contribution

The authors present a novel automated method integrating multiple techniques for robust detection and segmentation of liver and lymph node lesions, handling high lesion diversity.

Findings

Detection sensitivity of 0.53 for lesions.

Average Dice score of 0.71 for segmentation.

Robust handling of diverse lesion types.

Abstract

We propose a fully-automated method for accurate and robust detection and segmentation of potentially cancerous lesions found in the liver and in lymph nodes. The process is performed in three steps, including organ detection, lesion detection and lesion segmentation. Our method applies machine learning techniques such as marginal space learning and convolutional neural networks, as well as active contour models. The method proves to be robust in its handling of extremely high lesion diversity. We tested our method on volumetric computed tomography (CT) images, including 42 volumes containing liver lesions and 86 volumes containing 595 pathological lymph nodes. Preliminary results under 10-fold cross validation show that for both the liver lesions and the lymph nodes, a total detection sensitivity of 0.53 and average Dice score of $0.71 \pm 0.15$ for segmentation were obtained.