RESUMO
This paper presents a body-fixed-sensor-based approach to assess potential sleep apnea patients. A trial involving 15 patients at a sleep unit was undertaken. Vibration sounds were acquired from an accelerometer sensor fixed with a noninvasive mounting on the suprasternal notch of subjects resting in supine position. Respiratory, cardiac, and snoring components were extracted by means of digital signal processing techniques. Mainly, the following biomedical parameters used in new sleep apnea diagnosis strategies were calculated: heart rate, heart rate variability, sympathetic and parasympathetic activity, respiratory rate, snoring rate, pitch associated with snores, and airflow indirect quantification. These parameters were compared to those obtained by means of polysomnography and an accurate microphone. Results demonstrated the feasibility of implementing an accelerometry-based portable device as a simple and cost-effective solution for contributing to the screening of sleep apnea-hypopnea syndrome and other breathing disorders.
