Keywords:
Artificial Intelligence and Machine Learning, Artificial Intelligence, Lung, Musculoskeletal soft tissue, CAD, CT, Computer Applications-Detection, diagnosis, Segmentation, Chronic obstructive airways disease, Retrospective, Not applicable, Performed at one institution
Authors:
Z. Yang, H. Lee, T. Choi, J. Jung, M. Ryu, H. Yong; Seoul/KR
DOI:
10.26044/ecr2020/C-02697
Purpose
Chronic obstructive pulmonary disease (COPD) has been described as a systemic disease[1–6], which is known to be highly related to various adverse outcomes. The range of outcome is not only parenchymal function related disease, but also respiratory muscle-related diseases such as air trapping, dyspnea, reducing oxygen utilization, skeletal muscle dysfunction, and sarcopenia.
Pectoralis major muscle is one of the major respiratory muscle, covering most of the chest part and connecting the shoulder and abdomen. Many studies showed that the density and the volume of pectoralis muscles were highly related to COPD[7–16], by showing that the features of the respiratory muscle would be related in various parenchymal biomarkers, such as airflow limitation severity and diffusing capacity for carbon monoxide (DLCO)[17,18]. These are widely used features for diagnosing parenchymal diseases, such as asthma and COPD. The features of the pectoralis muscle have the potential to be used as an imaging biomarker, however, its measurements tended to be performed manually in specific slice location in CT volume with showing correlation with the volume of interest[7,8,10–13,19–21]. In-slice measured muscle areas may cause distortion due to muscle contraction and distraction from hold-breath level and posing while in CT scanning process.
To overcome this problem, deep learning-based automated volumetric pectoralis muscle segmentation method was designed for measuring the volumes and muscle mass for various application. Additionally, the preliminary result of the correlation between the former COPD diagnostic biomarkers and muscle volume/density, former various features, and basic radiomic features were calculated.