請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/53703
標題: | 雲端脈診分析系統以腎臟疾病為例 Web-based Pulse Analysis System for Detection of Kidney Disease |
作者: | Zih-Heng Wu 吳子衡 |
指導教授: | 賴飛羆(Fei-Pei Lai) |
關鍵字: | 電子化健康,醫療資訊系統,傳統中國醫學, eHealth,Medical information systems,Traditional Chinese Medicine, |
出版年 : | 2015 |
學位: | 碩士 |
摘要: | 傳統中國醫學歷史悠久且盛行於民間,近年來更興起中醫養身的風潮與學者們的關注,如何將中醫特有的把脈科學化、數據化使人信服並予人利用是中醫與西醫互相交流理解的首要課題。
臨床資料分析系統近年來在協助診斷上扮演著重要的角色,本研究之脈波分析系統旨在提供一雲端平台以供使用者可儲存、分享及分析病患脈波資料以達中醫科學化之目的。使用者可透過系統得到脈波資料經過快速傅立葉轉換、離散小波轉換、諧波及希爾伯特-黃轉換的結果,並使用這些結果進行統計及分類。 我們與國立臺灣大學醫學院附設醫院中的團隊合作,收集開心手術病人的脈波資料並予以分析,驗證此系統平台的可行性。該平台能累積脈波研究資料,並將資料分組分析。 本研究的實驗結果之一。發現有急性腎臟損傷的病人組在開心手術後的諧波二及諧波三較沒有急性腎臟損傷的病人組有大幅下降的趨勢並有顯著差異。另外我們也試圖透過脈波分析慢性腎臟疾病之特性。透過上述急性腎臟疾病的分析可知此系統或有臨床上鑑定急性腎臟疾病的潛力價值,但因為此實驗中急性腎臟疾病的病人數不多,仍須進一步的實驗與驗證。 Clinical data analytic system has played an important role in helping doctors for several years. The purpose of pulse analysis system is to provide a platform for researchers to analyze, share, and store patients’ pulse data. Researchers can use the system to extract features through Fast Fourier Transform (FFT), Discrete Wavelet Transform (DWT), harmonics, Ensemble Empirical Mode Decomposition (EEMD), and then take advantage of the former result to gather statistics and classify patients through two-sample t-test or paired-sample t-test. A group of researchers working in National Taiwan University Hospital (NTUH) have done an experiment to examine whether to find out acute kidney injury (AKI) through this system is feasible. The experimental subjects were patients who had received open heart surgery. The experiment result shows that after surgery the mean value of harmonics 2 and 3 of patients with postoperative AKI from the proposed pulse analysis system decreased significantly, but this does not happen on patients without AKI. In addition, we made an attempt to make a biomarker on chronic kidney disease (CKD) via pulse. This system would have a potential for clinical use to detect AKI after open heart surgery and probably other diseases. However, we need more cases to verify this experiment result because the number of AKI patients is limited. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/53703 |
全文授權: | 有償授權 |
顯示於系所單位: | 資訊網路與多媒體研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-104-1.pdf 目前未授權公開取用 | 1.17 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。