Optimizing Onset and Offset Detection in Surface Diaphragm Electromyographic Signals: A Signal Quality-Driven Approach.

This study explores the detection of onset and offset of diaphragm electromyographic activity in patients with chronic obstructive pulmonary disease (COPD) undergoing nocturnal non-invasive mechanical ventilation (NIV), aiming to enhance monitoring and management strategies for this patient population. The automatic analysis of electromyographic activity allows for the examination of respiratory patterns, addressing the challenges of cumbersome and time-consuming assessments. A signal quality-driven approach is used to improve the automatic detection algorithm for both onset and offset. The results show an improvement in the accuracy of onset and offset detection, with average onset detections advancing the scorer's mean by 194 ms, and offset detections lagging by 285 ms, which were reduced to 29 ms and 119 ms, respectively, after applying the correction algorithm. This approach has potential to enhance the monitoring and management of COPD patients undergoing nocturnal NIV.