Real-time apnea detection using pressure sensor and tri-axial accelerometer

Abstract

Respiratory disorders, if diagnosed late and untreated, may cause the advancement of many pathologies especially pertaining the cardiovascular system. This study proposes a method for a fast and certain detection of apnea events. For this purpose we used a commercial device that contains a pressure sensor helpful for the measurement of breath and a tri-axial accelerometer necessary to improve the detection reliability. There are numerous commercial devices able to detect breathing, but the totality of them is oriented to sport activity monitoring and so calibrated on the upper thresholds of respiratory rate. These devices are therefore not directly used as biomedical devices specific for the detection of sleep apneas or as life-saving devices in the case of “voluntary” apneas that occur in patients with severe neurological or pathological disorders. Then, we have developed complex algorithms to process the signals in real-time for the detection of apnea events with a maximum delay of 10 s, a sensitivity of 99%, and a specificity of 100%. This paper shows how an inexpensive approach is possible to control dependably the occurrence of apneas, avoiding hospitalization and the use of complex, invasive, and expensive devices.