Non-contact 3D measurement method for evaluating thoracoabdominal motion in respiratory assessment

Dysfunctional breathing causes negative impacts on health such as the myofascial pain syndromes and headaches. Biomechanical patterns of dysfunctional breathing are often assessed manually in clinical practice, and clinicians usually focus on thoracoabdominal motion created by the main respiratory muscles. We propose a non-contact method for measuring thoracoabdominal motion during breathing using a single depth sensor (Microsoft Azure Kinect DK) for automatic and quantitative biomechanics assessment of dysfunctional breathing. The measurement accuracy of the system is evaluated by comparing the data to that of the standard marker-based motion capture system (mocap). Ten healthy men participated in the experiment. The participants were asked to take deep breaths, and the thoracoabdominal motion was measured with the depth sensor and mocap. Reflective markers were attached to four landmarks on the body (just above the navel, xiphoid process, nipple line on the sternum, and sternum angle) for mocap measurement. The marker positions were estimated using depth images obtained by the depth sensor, and the errors in the mocap data were calculated. The standard deviation of the error in a breathing motion sequence was 2.2–4.7 mm. The proposed system can measure the thoracoabdominal motion with sufficient accuracy for clinical respiratory assessment.