Obstructive sleep apnea (OSA), as a highly prevalent sleep disorder, causes several serious health complaints. It has been proved that using intraoral mandibular advancement devices (MADs) during sleep is an efficient treatment for OSA. However, due to limited number of sleep study laboratories, effectiveness of MAD therapy is not regularly monitored. This paper proposes a smart MAD with the capability of continuously monitoring of cardiorespiratory parameters as well as sleeping postures and breathing routes. In this regard, a flexible hybrid wireless sensing platform based on the intraoral photoplethysmography (PPG), temperature and accelerometry monitoring is developed. It is qualitatively and quantitatively discussed that the intraorally captured PPG signals by the smart MAD have similar features as the ones received from the conventional anatomical position, i.e., the left index fingertip. Extensive experimental measurements indicate that the proposed smart MAD can estimate heart-rate (HR), respiration rate (RR), respiration pattern (RP), and blood oxygen saturation (SpO2) with an overall accuracy of 90% in comparison to the reference measurements, while such a capability is not dependent on subject's positions and breathing route. It is also shown that the smart MAD can readily identify different sleeping postures, namely, supine, left, right, and prone and breathing routes. The accuracy, reliability, repeatability, and stability of the proposed smart MAD's measurements are proved by examining a group of subjects. The proposed smart MAD has potential to monitor the effectiveness of MAD treatment and eliminate untreated OSA without the requirement of attaching an extra monitoring platform to the patient's body.

中文翻译：

---

智能下颌前移装置，用于心脏内呼吸参数和睡眠姿势的口内监测。

阻塞性睡眠呼吸暂停（OSA）是一种高度流行的睡眠障碍，会引起一些严重的健康问题。已经证明，在睡眠期间使用口腔下颌骨前移装置（MAD）是OSA的有效治疗方法。但是，由于睡眠研究实验室的数量有限，因此无法定期监测MAD治疗的有效性。本文提出了一种能够连续监测心肺参数以及睡眠姿势和呼吸路径的智能MAD。在这方面，基于口内光体积描记法（PPG），温度和加速度计监测的灵活的混合无线感测平台已被开发出来。定性和定量讨论了智能MAD口内捕获的PPG信号具有与从常规解剖位置即左食指接收到的特征相似的特征。广泛的实验测量结果表明，与参考测量值相比，拟议的智能MAD可以估算出心率（HR），呼吸频率（RR），呼吸模式（RP）和血氧饱和度（SpO2），总体精度为90％ ，但这种能力并不取决于受试者的位置和呼吸路径。还表明，智能MAD可以轻松识别不同的睡眠姿势，即仰卧，向左，向右以及俯卧和呼吸路线。拟议中的智能MAD'的准确性，可靠性，可重复性和稳定性 通过检查一组对象可以证明s的测量结果。拟议中的智能MAD有潜力监视MAD治疗的有效性并消除未经治疗的OSA，而无需在患者身上附加额外的监视平台。