Online Obstructive Sleep Apnea Detection on Medical Wearable Sensors

Obstructive Sleep Apnea (OSA) is one of the main under-diagnosed sleep disorder. It is an aggravating factor for several serious cardiovascular diseases, including stroke. There is, however, a lack of medical devices for long-term ambulatory monitoring of OSA since current systems are rather bulky, expensive, intrusive, and cannot be used for long-term monitoring in ambulatory settings. In this paper, we propose a wearable, accurate, and energy efficient system for monitoring obstructive sleep apnea on a long-term basis. As an embedded system for Internet of Things (IoT), it reduces the gap between home health-care and professional supervision. Our approach is based on monitoring the patient using a single-channel electrocardiogram (ECG) signal. We develop an efficient time-domain analysis to meet the stringent resources constraints of embedded systems to compute the sleep apnea score. Our system, for a publicly available database (PhysioNet Apnea-ECG), has a classification accuracy of up to 88.2% for our new online and patient-specific analysis, which takes the distinct profile of each patient into account. While accurate, our approach is also energy efficient and can achieve a battery lifetime of 46 days for continuous screening of OSA.

A Modular Approach towards Wearable Computing

Xavier Righetti

Thanks to recent improvements in electronics, conductive materials and energy storage capabilities, we are now able to embed small circuits but also ultra-mobile PCs (UMPC) into clothes. These devices are able to sense, process and communicate information between the user and his/her context. During the first research phase, we have developed a set of wearable applications that aimed at assisting the user in various situations such as the teleoperation of a blimp, or the pedestrian navigation in an unknown indoor environment. Research on both of these experiments can be classified in three categories: First the multimodal interaction between the users, his/her wearable interface, and the environment. Second, the context sensing such as the location of the user. Third, the wireless communication between the embedded system and the surrounding or remote devices. Although the resulting functionalities were satisfying as expected, both systems implied the use of cumbersome equipment such as a Head-Mounted Display (HMD), or a UMPC, and thus were too restrictive. With the experience gained from these experiments, we have decided to pursue a different approach. Indeed, in order to develop a truly wearable system, we have focused our research on the design and development of wearable modules based on small 8-bit microcontrollers. In our vision, one module equals one functionality. The advantages are threefold: First, the applications are virtually unlimited as the wearable modules can be attached or removed on the fly. Second, the user only wears what is needed, thus eliminating the superfluous hardware. Third, compared to commercial wearable products, the user can change the garment without changing the functionalities, or on the opposite, s/he can change the functionalities without changing the garment. A special attention has been put in the design of magnetic connectors in order to facilitate the placement of the modules on the garment. Compared to alternative approaches such as the lilypad from Leah Buechley, our modules do not need to be sewn to the garment and thus can very easily be relocated or removed when washing the garment for example. The decision to opt for a wired bus has been motivated by several factors. Indeed, not only the modules consume less power, but also they are smaller and lighter as they don't need wireless functionalities, nor individual batteries. We have developed specific applications in order to demonstrate and evaluate our modular approach. The feedback collected from test users and the results of various evaluation techniques showed us that our system is comfortable, attractive and perceived as flexible and creative. We believe that this kind of approach may not only foster the involvement of the user in wearable computing applications, but it can also be seen as a rapid prototyping platform consisting of a set of tools that facilitate the development of wearable applications.

EPFL2010

Real-Time Probabilistic Heart Beat Classification and Correction for Embedded Systems

David Atienza Alonso, Srinivasan Murali, Francisco Javier Rincon Vallejos, Grégoire Casimir Joseph Surrel

With the emergence of wearable and non-intrusive med- ical devices, one major challenge is the real-time analysis of the acquired signals in real-life and ambulatory con- ditions. This paper presents a lightweight algorithm for on-line heart beat classification and correction that relies on a probabilistic model to determine whether a heart beat is likely to happen under certain timing conditions or not. It can quickly decide if a beat is occurring at an expected time or if there is a problem in the series (e.g., a skipped, an extra or a misplaced beat). If an error is detected, the series is repaired accordingly. The algorithm has been carefully optimized to minimize the required processing power and memory usage in order to enable its real-time embedded implementation on a wearable sensing device. Our experimental results, based on the PhysioNet Fanta- sia database, show that the proposed algorithm achieves 99.5% sensitivity in the detection and correction of erro- neous beats. In addition, it features a fast response time when the activity level of the user changes, thus enabling its use in situations where the heart rate quickly changes.

2015

A Modular Approach towards Wearable Computing

Xavier Righetti

EPFL2010