Investigating the Effectiveness of Mandibular Advancement Devices and Maxillomandibular Advancement Surgery for the Treatment of Obstructive Sleep Apnea

Gruenspan, Grant 1, 2 ;  Saha, Shumit 1, 2 ;  Meisami,Tina 2, 3 ;  Yadollahi, Azadeh 1, 2

1. Institute of Biomaterials and Biomedical Engineering, University of Toronto, Canada; 2. Toronto Rehabilitation Institute-University Health Network, Canada; 3. Faculty of Dentistry, University of Toronto, Canada

Background: Obstructive sleep apnea (OSA) occurs due to the collapse of pharyngeal airway during sleep. The severity of OSA is assessed by the number of partial (hypopnea) or complete (apnea) collapses in the pharyngeal airway per hour of sleep (apnea-hypopnea index (AHI)). Mandibular advancement with oral appliances and maxillomandibular advancement (MMA) surgery are popular treatments for OSA which involves opening up the pharyngeal airway. However, these methods are not effective in all OSA patients, and there is no objective method to predict the effectiveness of these treatments. Previous studies have shown that structural changes to the pharyngeal airway can affect the craniofacial structure and generation of the breathing and speech sounds. Thus, we aim to investigate the feasibility of craniofacial and sound features to predict the effectiveness of mandibular advancement devices and MMA surgery.

Method: In this study, 30 OSA patients who will undergo mandibular advancement or MMA surgery at the oral surgery clinic located at Toronto Rehabilitation Institute will be included. We will take photographs of the head and neck to analyse the craniofacial structure. Breathing and vowel sounds will be recorded by two microphones. The pharyngeal airway length and area will be measured by acoustic pharyngometry. Before and after MMA surgery or applying a mandibular advancement device, all measurements will be repeated. Before and after the treatment, the AHI will be assessed using a wearable monitoring device. If the AHI doesn’t reduce after treatment, then the patient will be classified as a non-responder to the treatment. From the photographs and the recorded sounds, the craniofacial and acoustic features will be extracted, respectively. Subsequently, we will assess the changes in craniofacial and sound features before and after the treatment in both the non-responder and responder group. Based on the assessment, we will build a feature set and a prediction model for estimating the effectiveness of these treatments.

Expected Results: Previous studies have shown that OSA patients with larger mandible width and face width may have a higher AHI, which may lead to a less effective treatment with mandibular advancement. Moreover, previous studies have shown that reduction in the pharyngeal airway is associated with the louder sound intensity. Therefore, we expect to extract features such as mandible width, face width and sound intensity, which will have the ability to predict the effectiveness of the mandibular advancement or MMA surgery. Thus, combining these features will make a screening technology for assessing the effectiveness of these OSA treatments.

Significance:  The screening technology will help physicians to determine the subgroup of OSA population who will respond significantly to the mandibular advancement and MMA treatment. This will significantly reduce the cost for the patient, and help them to get an optimum treatment.