穿戴式生理感測裝置之整合應用平台

An Integrated Wearable Application Platform with Physiological Sensors

邢維翰(Hsing, Wei Han)

指導教授 ： 馬席彬

國立清華大學/電機資訊學院/通訊工程研究所/碩士(2015年)

摘要

在論文中，我們提出以穿戴式生理感測裝置為基礎，整合出相關的應用平台。 我們使用穿戴式裝置收集心電訊號和呼吸訊號，第一種穿戴式裝置是由前端電路、FPGA控制器和藍芽模組所構成的。另一種穿戴式裝置是使用三軸加速度計和麥克風來收集呼吸訊號和鼾聲音訊。我們將訊號處理在手機上實踐，並在Android系統中設計app，方便使用者操作。
為了更快速、方便地分析生理訊號，我們選擇雲端系統來做為計算的平台，我們使用的雲端系統為Apache Storm，可以將資料透過無線網路上傳至雲端伺服器。藉由這些雲端計算的優點，我們提出一個有關情緒辨識的應用，透過穿戴式裝置收集生理訊號並整合雲端計算，完成訊號分析的工作。我們選擇的分類器:支持向量機，可將不同種類的訊號作分類，最終的分類準確率為58.3%。
除了以上的應用，我們也將穿戴式裝置延伸至其他應用。心腦互動的應用是透過心電訊號及腦波訊號得知手術前後的生理差異，並依據其差異判別使用者的情緒反應，特別使用顏色來區分緊張與放鬆的情緒差異。睡眠呼吸障礙患者的監測平台是利用三軸加速度計和麥克風收集生理訊號，將手機當作監測的平台，觀察患者的呼吸狀態及判定呼吸中止症的發生。最後我們提出一個結合呼吸訊號與娛樂的應用。

關鍵字

穿戴式裝置

並列摘要

In this thesis, an integrated wearable application platform with physiological sensors are presented. We use the wearable devices to collect the electrocardiography (ECG) respiration (RESP) signals. The first wearable device is a prototype with analog front-end, FPGA controller and Bluetooth module. The other wearable device is using accelerometers and a microphone to collect respiration signals and snore sound. The mobile phone is a platform for dealing with the digital signal processing. We design an Android app with convenient user interface for every application.
In order to get more convenient to analyse these biomedical signals, we choose cloud computing system to be our computing platform. We implement the cloud computing by Apache Storm, which can transmit the data by streaming via 3G/Wi-Fi. Depending on the advantages of the cloud computing, we develop an application with the wearable device for the emotion recognition. The extracted-features of biomedical signals are implemented with methods by the cloud computing. The support vector machine (SVM) is a classifier for classifying the data set into the correct classes. The accuracy of the classification is 58.3 %.
Besides the detecting emotion recognition system, this wearable device can be applied in various applications. The heart and brain crosstalk system is an application of using ECG and Electroencephalography (EEG) signals to observe the patients’ surgery recovery. The special part is that the mental status of the patients can use colors scale to represent. The home monitoring system for sleep-disordered breathing is an application to monitor the patients
who is obstructive sleep apnea (OSA) and uses the accelerometers and microphones to detect the symptoms of disease. We modify the monitor platform from using MATLAB to Android mobile phone. The last application is using the same wearable device to collect RESP signals and then develop an entertaining app game.

並列關鍵字

wearable device

參考文獻

[1] W. B. Cannon, Bodily changes in pain, hunger, fear and rage, an account of recent researches into the function of emotional excitement. New York and London, D. Appleton

and Company, 1915.

[4] H. Holzapfel and C. Fuegen, “Integrating emotional cues into a framework for dialogue management,” in Proc. 4th IEEE Int. Conf. Multimodal Interfaces. Washington, DC,

USA: IEEE Computer Society, 2002, p. 141.

[5] C. Peter and A. Herbon, “Emotion representation and physiology assignments in digital systems,” Interacting with Computers, vol. 18, no. 2, pp. 139 – 170, 2006.