Computerized acoustic detection of obstructive apnea.

Cardiac, respiratory and neurologic abnormalities have been identified as causes of Sudden Infant Death Syndrome (SIDS). Recurrent central apnea (no respiratory effort or nasal/oral airflow) and obstructive apnea (respiratory effort without concurrent nasal/oral airflow) in infants are considered risk factors for SIDS. However, using currently available monitoring techniques, normal activities such as yawns, stretches and swallows cannot be distinquished from short obstructive episodes lasting less than 20 s. A system was developed to more accurately detect obstructive apnea in infants using a miniature microphone placed over the trachea, a cassette tape recorder and a MINC-11 microcomputer. Respiratory sounds were recorded on 5 anesthesized rabbits in which partial and total airway obstruction was artificially induced. Sounds were analyzed by computer using fast Fourier transformations. Amplitude versus frequency was plotted for normal breathing, partial obstruction and total obstruction. Characteristic patterns were identified for each episode demonstrating that acoustic detection of apnea in infants by a microprocessor-based monitor is feasible.