| 摘要: | 近年來，健康照護與醫療復健逐漸受到重視，加上科技日新月異，許多傳統的醫療設備在結合新的電子技術後，使其產生了更強大的功能。在醫療復健上，因為許多的醫療設備尚未與電子技術結合，使得醫師無法隨時掌握病人復健的情況，而病人在家裡進行復健時，因為沒有醫療人員的協助下，常無法得知自己的動作是否正確，進而影響復健的成效。
在本篇論文中，我們提出一新的醫療架構，將傳統的醫療復健設備與電子技術結合，再透過資料的蒐集、量化、分群、分類與分析等步驟，來協助醫生進行病人的復健，其中，經由病人復健資料的分群，我們學習了病人的復健動作，再透過分類的過程，來比對病人的復健動作是否正確，以幫助病人自行在家中進行復健。
實作上，我們將此架構應用於醫學上的骨盆底肌肉訓練，我們與台北榮總婦產科的醫師合作，將系統實際應用於尿失禁的患者上，並用實驗的數據，來證明系統能有效的協助病人進行復健，進而改善他們的復健療程。
Recently, healthcare and rehabilitation are becoming more and more important. Due to the advancement of technology, traditional medical equipment may produce great contributions after the combination of equipment and electronic technology. Nowadays, much medical equipment is not combined with electronic technology. Hence, doctors do not track patients’ courses of treatment. Moreover, when patients do rehabilitation exercises at home without the assistance of doctors or nurses, they cannot confirm that their rehabilitation motions are correct or not. It may affect the effectiveness of rehabilitation.
In this dissertation, we propose a new rehabilitation architecture. We connect traditional medical equipment with electronic technology. We assist doctors in understanding patients’ rehabilitation via the process of data collection, quantization, clustering, classification, and analysis. Also, by the process of clustering and classification, patients can determine whether their rehabilitation actions are correct or not and adjust their own rehabilitation actions accordingly to improve the effectiveness of rehabilitation.
In implementation, we apply our architecture to pelvic floor muscle training (PFMT). We cooperated with a doctor of the Department of Obstetrics and Gynecology, Taipei Veterans General Hospital, Taiwan, and performed a series of experiments to verify our theory. The system has been implemented in medicine and has been used to treat patients with urinary incontinence in practice. We hope that this system can help more patients in the future. |
