Sleep studies: going wireless?

Caren B. Les, caren.les@laurin.com

Those suffering from insomnia may be less inclined to have their sleep clinically evaluated because of the numerous wired devices to which they may have to be attached. Scientists have recently tested the accuracy of thermal infrared imaging as a noncontact method of monitoring a patient’s airflow during polysomnography, a comprehensive test for studying sleep and diagnosing sleep problems in humans.

Airflow monitoring is used to evaluate patients for signs of obstructed breathing disorders during sleep, such as apnea and hypopnea. Polysomnography typically involves the measurement of several bodily functions, including eye movements, heart rhythm, and brain and muscle activity. A patient can be attached to at least 20 contact sensors and associated wires during a sleep study, which for diagnosing sleep apnea may include nasal prongs and oro-nasal thermistors, and the more intrusive reference standard airflow sensor, the pneumotachometer. This apparatus can cause discomfort to a patient, and contact sensors may misalign during sleep, which can affect data quality and requires repositioning by a technician.

Dr. Jayasimha N. Murthy, assistant professor of medicine in the division of pulmonary, critical care and sleep medicine at the University of Texas Health Science Center in Houston, and Ioannis Pavlidis, director of the computational physiology lab at the University of Houston, contributed to the investigation in the areas of medical aspects and applications, and engineering aspects, respectively. Their scientific report was accepted by the journal Sleep.

They demonstrated the feasibility of using thermal infrared imaging to remotely measure breathing rates of a sleeping patient. The method was based on the second order statistics of the presence or absence of hot expiratory plume in the area of the nostrils. The investigators also reported a more advanced method of wavelet analysis of the thermal imaging signal on the nostrils themselves. They were able to extract the full breathing waveform (rate and amplitude). Through consistent segmentation of the nostril area and facial tissue-tracking algorithms, they sustained monitoring of breathing that was functionally equivalent to the thermistor method. They found that, during monitoring of healthy individuals, agreement between breathing waveforms from the virtual (imaging) and contact thermistors was greater than 90 percent.

Known as the Automatic Thermal Monitoring System (Athemos), their method consists of a mid-wave IR camera, a blackbody for calibration and a pan-tilt mechanism for positioning. The team developed customized software to control the system and to process the acquired imaging signals in real time.

Study participants consisted of 14 volunteers with no history of sleep disordered breathing and 13 patients with a history of sleep apnea. The subjects underwent polysomnography for one to two hours, during which recordings from respiratory plethysmography, conventional airflow channels, thermal infrared imaging and other physical functions were obtained. The investigators determined that for detecting apnea and hypopnea, the noncontact thermal infrared imaging method is comparable in accuracy to the conventional flow sensor methods of thermistor, nasal pressure and continuous expired CO₂ waveform methods.

The investigators noted that the method must be validated with further studies to evaluate its potential in clinical sleep medicine.