建構高速無線感測網路平台用於遠距醫療照顧系統

Abstract

老人化社會的來臨，對現行醫療服務帶來巨大的挑戰。特別是在醫院中，面對老人化社會帶來的大量醫療服務需求，傳統醫療型態會對醫療人力帶來更大的工作負荷。例如，護士的工作負荷將會隨老人病患的增加而加重。面對此挑戰，近年來廣泛被研究的遠距醫療照護系統即是一個解決方案。遠距醫療照護系統能監測病人的生理資訊，並且在危急時透過無線網路通知醫院中的護士站。 回顧過去之研究，前人廣泛地使用ZigBee無線感測網路來架構遠距醫療照護系統；然而，卻僅止於討論如何製作硬體平台，或是建立系統功能。ZigBee的最大缺點：低資料量沒有被考量在系統之中。此缺點會直接影響網路的節點容量，以及實際能運作的節點數量。本研究對此提出了解決方案，實作了一款以數位訊號處利器為主的ZigBee無線感測網路平台，並且使用此平台發展了一套醫院用之遠距醫療照顧系統。除此之外，本研究改良了一套以小波轉換為基礎的心電圖壓縮演算法，提供更佳的壓縮效能，並且實作於平台上，有效地降低網路中的資料量並提昇網路的有效節點容量。 在結果部份，本研究使用Markov模型分析了ZigBee網路中的有效資料量，並且透過實驗來驗證。此外，ZigBee的基本特性，如封包遺失率、訊號強度變化等，也由實測取得數據，方便未來研究參考。在演算法的部份，本研究比較了改良前後的心電圖壓縮演算法之效能差距，並且整合了判別人體動作之演算法於系統之中。與其它的系統相較，本系統以節省頻寬與降低網路資料流量的方式，將高資料量的醫院用遠距醫療系統，使用低資料量的ZigBee無線感測網路實現。進一步而言，結合數位訊號處理器平台與壓縮演算法，未來使用低資料量的ZigBee系統將可實現更多的高資料量應用。

Aging society brings great challenge to nowadays healthy service. Especially in hospital, the traditional medical service style which wasted the manpower of the healthy professionals shall be changed to handle the burst of the regular physical examination from the elder patients. For example, the working load of the nurses will be increased since the elderly needs more physical examinations. As a solution, a telecare system can monitor personal physiological status and can give alarm to the nurse station in a hospital through wireless network. ZigBee, as a wireless standard, was used by previous research to build such system. However, previous research which built the telecare system by ZigBee did not consider the data rate problem but only mentioned how to build a platform or how did the telecare system function. Although these research claimed that ZigBee brings a chance to build a multiple nodes wireless sensor network, but the data rate problem shows counterevidence of it since the problem affects the capacity of the network. In this study, a ZigBee telecare system and a new DSP-based platform are proposed. A modification of an ECG compression algorithm is also implemented and examined in the system and successfully reduces the redundant data of bio-signals. The proposed telecare system can provide ECG and posture monitoring service with low network traffic. In result, this study uses simulations and experiments to analysis the effective data rate of ZigBee. Moreover, other basic performance such as packet lost rate and received signal strength index were evaluated in order to obtain some reference values for future ZigBee-based application. The result also shows the performance comparison between modified compression algorithm and unmodified one. In conclusion, the proposed system offers lower packet lost rate and lower network traffic than the telecare system without compression algorithm. Furthermore, combining the proposed platform and the compression algorithm, the chance of implementing a high data rate application on a low data rate ZigBee system is bloomed and waiting to be glorified.