Characterization Of Chimeric Surface Submentalis EMG Activity During Hypopneas In Obstructive Sleep Apnea Patients MAK A. Daulatzai, Ahsan H. Khandoker, Chandan K. Karmakar, Marimuthu Palaniswami Sleep Disorder Group, Dept. of EEE The University of Melbourne Melbourne, Australia makd@unimelb.edu.au Neela Khan Faculty of Life and Social Sciences Swinburne University Melbourne, Australia Abstract— Polysomnogram (PSG) is the standard diagnostic test for the evaluation of sleep disorders. The current rules require surface (s) electromyography (EMG) of the submentalis muscle (SM) in order to document atonia during REM sleep. The sSM EMG signals reflect contracting motor units; the firing of the latter is a function of intrinsic neuromuscular characteristics of its component muscle fibers, and indeed forms the basis for the spectral properties. Here we have studied OSA patients with apnea-hypopnea index (AHI) of <5, 5-10, 30- 35, and 60+, and document, for the first time, a “Chimeric” sSM EMG activity phenotype during hypopneas in Non-REM sleep. This unique pattern characteristically displays contiguous tonic-phasic segments of high activity and low activity or vice versa. We have analyzed the total duration, and other attributes of these hybrids in comparison with the normal awake and apnea/hypopnea-free sleep periods. We document an inherent heterogeneity between hypopneas, and between heterogeneous segments of the chimeras in OSA patients of varying AHI. This study emphasizes that rectified and filtered sSM EMG activity signals provide a novel, valid and useful metric in PSG evaluation which may be of clinical significance in sleep-related and other pathological conditions. Keywords- polysomnogram, surface electromyography, submentalis muscle, apnea-hypopnea index, Chimeras, hybrid activity, novel metric, sleep diseases I. INTRODUCTION Obstructive sleep apnea (OSA) is a major risk factor for a number of clinical conditions. These include cardiovascular conditions including hypertension [1-3], myocardial infarction [4, 5], neurocognitive impairment [6], stroke and sudden death [7, 8]. It is also linked to various other diseases such as memory perturbations, depression, and poor cognition performance [9]. Sleep disordered breathing ranges from snoring through increased airway resistance and reduction in airflow (hypopnea) to periods of breathing cessation and airway collapse (apnea). Apnea-hypopnea index (AHI) is the average number of apnea and hypopnea per hour of sleep. The AHI number is taken to characterize the severity of OSA. An AHI of >5 is generally regarded as abnormal. Mild disease is reflected by 5-15 AHI, moderate by 15-30, and severe by an AHI greater than 30 per hour. UA negative pressure is the critical factor, and the UA occlusion during sleep is caused by the subatmospheric pharyngeal pressure during respiration, and reduced UA dilator muscle activity [10-13]. OSA causes hypoxemia, multiple arousals (resulting in daytime sleepiness) increased respiratory effort, sleep fragmentation, and various pathophysiological sequelae [14]. The classical epidemiological data have characterized hypopnea utilizing thermisters and/or inductance plethysmography in association with a 4% oxygen desaturation. However, more recently nasal pressure transducers are used in scoring hypopnea. There are several clinical definitions of hypopnea in use, and hence different labs use different criteria owing to a lack of consensus on this issue [15]. Although hypopnea and apnea are not equal and differ significantly, the physiological outcome of hypopneas have been considered in various studies to be similar to that of apnea. Despite the fact that consequences of hypopnea in low AHI subjects may seem not to translate in significant pathology, in the long run however, they are devastating since it is the progression in quality and quantity of hypopneas that eventually results in the genesis of apnea and OSA-associated co-morbidities. As such studies that may throw any light on different facets of hypopnea are called for. An important component of polysomnograms (PSG) is surface submental electromyography (sSM EMG). This is a non-invasive method that depicts electrical activity from submental as well as from the muscles in its close proximity. sEMG sensors are placed in and around the submental area between the mandible and the hyoid bone, and as such measures muscles of the mouth floor [16]. Although sSM EMG has been extensively studied in dysphagic patients, there TIC-STH 2009 978-1-4244-3878-5/09/$26.00 ©2009 IEEE 782