具定位服務之手持式RFID醫療輔助系統之研究

Abstract

在醫院中，各項醫療資訊散佈於院內各處，這些資料可能會在不同的地點、不同時間，透過不同管道的與不同對象進行傳輸。因此，如何有效地進行資訊傳遞，並提昇醫護人員行動性，就成了重要的議題。 隨著行動運算技術的成熟與無線網路的普及，醫院中開始導入行動醫療輔助系統來提昇院內資訊傳輸效率，並改善病人安全。在2006年，我們與萬芳醫院合作，開發了一套「手持式RFID醫療輔助系統」供院內住院病房區使用。系統結合了行動裝置、Wi-Fi無線網路、RFID、Web Service等技術，提供包括：行動醫療資訊平台，減少醫護人員來回護理站的次數。高風險查驗檢警示系統，漸少檢查檢驗資料流程的延遲。RFID患者身份辨識機制，改善傳統利用條碼辨識的造成不便。並結合線上藥典顯示用藥安全資訊等服務，藉此改善院內醫療服務品質。 此外，我們還在原本的系統架構中結合了位置感知系統，提出了「位置感知資料下載」與「位置感知行動資料庫」兩種新的機制，減少醫護人員使用行動裝置與無線網路與遠端進行資料傳輸而造成的延遲的機會。其中，行動裝置中的「位置感知行動資料庫」會週期性地依照系統預測之行動裝置位置，進行「位置感知資料下載」，與醫療資料庫進行資料同步，確保資料的即時性。而醫護人員在理想狀況下，可以直接由行動裝置中的「位置感知行動資料庫」，存取所在區域即時的醫療資訊，取代以往由遠端資料庫下載的方式。而根據我們的實驗結果發現，導入位置感知機制後的資料存取時間，比起傳統由遠端下載的方式改善了64%。 而要提供位置感知服務，必須結合一套室內定位系統，而本研究也提出了一種以Fingerprinting演算法為基礎的手持式Wi-Fi室內定位機制。根據定位實驗結果，可以達到平均兩公尺左右的誤差，並具有將近百分之九十的區域辨識成功率。 此外，結合Wi-Fi定位系統還可以週期性監控行動裝置的即時狀態，並讓醫護人員隨時查詢目前院內行動裝置使用者以及位置等資訊，幫助院方管理行動設備使用狀況。 最後，介紹了六個國內外導入行動醫療輔助系統的案例，並且與本研究所提出之系統進行多方面的比較。結果顯示，與國內系統相較，在位置感知服務以及異常報告之項目上成為本系統之優勢。而系統弱勢的部份，則在於醫藥囑開立以及手術資訊整合的部份，未來本系統則仍會針對這些缺失持續做改進。而與國外系統相較後，則發現本系統對於位置感知服務系統的整合度以及功能尚不若國外系統來得完善，因此也將繼續提昇位置感知服務的品質，並加強前後端系統的整合。

Plenty of healthy records and data are scattered over the hospital. Healthy information may be delivered to different targets by different ways anytime and anywhere in the hospital. So, the problem of how to transmit data efficiently and increase the mobility of personnel is taken seriously. With the advancing and popularizing of mobile computing and Wi-Fi wireless network, some hospitals are trying to introduce mobile medical assistant system to improve the data transmission mechanism and patient safety. In year 2006, we cooperated with Wan Fang Hospital and then developed a “Handheld RFID Medical Assistant System” for ward area. It was combined with many technologies such as mobile device, Wi-Fi, RFID, Web service to provide a mobile medical information platform to reduce the chance that personnel go back to the nurse station for medical data, high risk reminder mechanism to improve the time-wasted procedure of laboratory test, emergent RFID against disadvantages of traditional barcode identification and a combination with online pharmacopoeia to show medication safety information at bedside. We hope this system can raise the quality of healthcare. Besides, in this research, we add a new location-aware mechanism to the original system and carry out two ideals of “location-aware data transmission” and “location-aware mobile database” to reduce the risk of remote data accessing delay when using mobile device. The location-aware database in mobile device lunches location-aware data transmission process periodically to synchronize local-area data with back-end hospital information database according to the estimated location itself. In the ideal condition, personnel can access the local data from the location-aware database in their mobile instead of the time-consumed remote communication with back-end database. According to our experiment result, the new mechanism makes an improvement of 64% in data transmission time. However, we need an indoor position system to provide estimated location to the location-aware service module. A handheld Wi-Fi location system which follows the fingerprinting algorithm is also proposed in this research. According to the result of our indoor positioning experiments, we have an average error about 2 meters and nearly 90% correctness of region identification. With the help of this position system, we can also locate the online mobile device in the hospital and help personnel to get over the inconvenient for mobile device management. Finally, we list six case studies of other proposed mobile medical systems and make a comparison with ours. According to the comparison, our system has better location-aware and high risk reminding mechanism than other system in Taiwan. But the integration with medical order and operation information of our system are weaker and we will take these weaknesses as future works. Besides, by the comparison with those in other countries, we find that their location-aware services are better than ours in quality and degree of integration. So we will also continuously improve our location-aware service and make a stronger combination between front-end and back-end systems.