Automated segmentation of muscle and adipose tissue on CT images for human body composition analysis

Abstract

The ability to compute body composition in cancer patients lends itself to determining the specific clinical outcomes associated with fat and lean tissue stores. For example, a wasting syndrome of advanced disease associates with shortened survival. Moreover, certain tissue compartments represent sites for drug distribution and are likely determinants of chemotherapy efficacy and toxicity. CT images are abundant, but these cannot be fully exploited unless there exist practical and fast approaches for tissue quantification. Here we propose a fully automated method for segmenting muscle, visceral and subcutaneous adipose tissues, taking the approach of shape modeling for the analysis of skeletal muscle. Muscle shape is represented using PCA encoded Free Form Deformations with respect to a mean shape. The shape model is learned from manually segmented images and used in conjunction with a tissue appearance prior. VAT and SAT are segmented based on the final deformed muscle shape. In comparing the automatic and manual methods, coefficients of variation (COV) (1 − 2%), were similar to or smaller than inter-and intra-observer COVs reported for manual segmentation.

IODA: An input/output deep architecture for image labeling

“…To the best of our knowledge, the only automated method for skeletal muscle segmentation at L3 was proposed by Chung et al [5]. The method is based on standard shape prior coupled with an appearance model.…”

“…We evaluated the proposed IODA automatic segmentation in comparison with the manual segmentation, and also with a referenced method [5] proposed by Chung et al briefly described below. In order to evaluate and compare their performance, the Jaccard index was used to measure the overlap between IODA (respectively Chung's) segmentation and the manual segmentation.…”

“…The originality of IODA is to transpose DNN pre-training input trick to outputs, in order to learn a high level representation of labels. We apply it on a medical imaging labeling problem on which we outperfom the state-ofthe-art method achieved by a dedicated approach based on an a registration on an a priori model [5].…”

Image Segmentation for Cardiovascular Biomedical Applications at Different Scales

“…Recent research on influence of different tissues [34] on ECG signals suggests that muscle, lungs and fat tissues are more important than the others. Most of these tissues can be segmented automatically [35][36][37][38]. In this paper, we will focus on segmentation of abdominal parenchymal organs.…”

Paraspinal muscle segmentation in CT images using a single atlas

“…It has been used to segmentation abdominal organs [20], brain tissues [21][22], and heart images [23]. Atlas-based segmentation combined with PCA encoded Free Form Deformations has been applied to segmenting body muscle on CT images for human body composition estimation [24]. Global image registration and local contour optimization have been used to segment abdominal organs on images belonging to a sequence of CT images [25].…”