脈診是今日中醫最熱門的診斷方式之一，但是由於缺乏量化分析實驗與臨床數據，中醫在許多國家仍然被視為一種補充或替代醫學，因此要改善中醫目前的困境，將中醫科學化必須把數千年以來的理論與診療方式用精確的文字或數據描述。雖然目前已有不少脈診儀器問世了，但卻是以單維的脈搏波為主，只有本團隊擷取脈波影像，然而在自動取脈方面仍缺乏即時迴饋，而無法準確地將感測器放在脈管中央。另外在脈波影像分析也缺乏臨床資料的分析。

為了要量化分析在成功大學附設醫院蒐集病歷的脈波影像，在本論文中沿用了兩種分析方法，等高線分析與L-cube方程式分析。等高線分析主要是針對時間與空間兩種不同維度交互作用的分析，利用設定不同閥值在二維脈搏波圖上繪出相對應等高線，然後利用學生T-test進行統計檢定。L-cube方程式分析則是利用方程式擬合三維脈搏波圖譜，將擬合最佳化後的參數作為特徵值，藉由支援向量機進行分析。本論文設計了一實驗利用上述兩種方法針對具有以及不具有心血管疾病高風險因子兩個族群進行檢驗，用以驗證此兩方法的可行性。結果顯示在閥值設定為0.8時的等高線進行分析，二維長軸（2DL）、二為橫軸（2DT）與兩軸比（2DL/2DT），在統計上皆呈現有顯著差異；另外利用L-cube方程式提取特徵值進行分群測試也有83.2%的正確率。

為了改善無法即時回饋取脈的問題，本文提出的即時回饋之自動取脈系統包含改良的即時濾波器，有窗口且加權過之斜率和方程式以及脈波波峰偵測演算法，可讓脈波影像儀將機械手指正確放置在波峰頂點位，進而增加脈波影像儀的操作效率，有了這一系統後可使取脈流程更加快速而且精確，可望加速中醫的量化分析。

The pulse diagnosis is one of the most popular diagnosing method in Chinese medicine nowadays, but it is still treated as a kind of complementary and alternative medicine in many countries because of the leakage of quantitative analysis. In order to improve the situation of Chinese medicine, describing Chinese medicine theory and the process of diagnosis in quantitative is the first step of scientific research in Chinese medicine. Although, there are many instruments have been published, those instruments record mostly in one dimension. Only the instrument which invented by our team records pulse image. The automated pulse-taking process still leakages of real-time feedback, thus the actuators cannot place sensors right on the center of radial arteries precisely. Furthermore, the clinical sample is not enough for analyzing pulse image.

In order to quantifying pulses, two analyzing methods are proposed in this thesis. The contour analysis focuses on the both time domain and space domain. It weights different threshold on 2DPM to plot contour line, and adopting student T-test for analyzing. L-cube formula analysis quantifies the pulses by using a formula to fit the 3DPM, and taking the critical parameters of L-cube as features for support vector machine (SVM) to learn.

For the sake of testing the performance of two methods, the experiment is designed to classify the population with and without high cardiovascular risk. The result of contour analysis shows significant difference in statistical while weighting of contour lines at 0.8 no matter 2DL ,2DT or the ratio of 2DL to 2DT. The result of L-cube formula analysis shows that there is 83.2% accuracy to classify two groups.

In order to improve the leakage of real-time feedback in automated pulse-taking process, this thesis proposed the automated pulse-taking system consists of an improved real-time filter, a window and weighted slope sum function and an algorithm which evaluating the pulse peaks. It can accurate locating the artificial finger onto the peak spot. The automated pulse-taking system makes the operator take pulses more efficiency. PII (Pulse Image Instrument) with automated pulse-taking protocol makes data acquisition more precisely and efficiency than previous version. The improvement hopes to enhance the progression of quantifying Chinese medicine.

[1] World Health Organization. "WHO international standard terminologies on traditional medicine in the western pacific region." (2007).
[2] 佚名, “難經”, B.C. 206 – B.C. 220
[3] 張景嶽. "景嶽全書." 太原: 山西科學技術出版社 389 (2006).
[4] 孟景春, 周仲瑛. "中醫學概論." 台北: 知音 (1994).
[5] 王敬義, “脈論: 二十年後方為醫.” 北京: 中醫古籍出版社, 2009.
[6] Luo, Ching-Hsing, et al. "Possibility of quantifying TCM finger-reading sensations: I. Bi-sensing pulse diagnosis instrument." European Journal of Integrative Medicine 4.3 (2012): e255-e262.
[7] N. Yu, “Pulse-taking Protocol Study and Wrist-fixer Feedback Control of Pulse Mapping Instrument for Chinese Western Medicine,” 2016.
[8] Luo, Ching-Hsing, et al. "Non-invasive holistic health measurements using pulse diagnosis: I. Validation by three-dimensional pulse mapping." European Journal of Integrative Medicine 8.6 (2016): 921-925.
[9] Su, Chun-Jen, Ting-Yi Huang, and Ching-Hsing Luo. "Arterial Pulse Analysis of Multiple Dimension Pulse Mapping by Local Cold Stimulation for Arterial Stiffness." IEEE Sensors Journal 16.23 (2016): 8288-8294.
[10] C.S. Hu, “Three Position Nine Indicators Pulse Diagnosis Robust Study in Traditional Chinese Medicine Using Pressure Sensor Array,” 2011.
[11] C.W. Yeh, “Automation and Measurement Protocol Design of Second Generation Bi-Sensing Pulse Diagnosis Instrument for Chinese and Western Medicine,” 2015.
[12] C.H. Chen, “Implementation of Second Generation Bi-Sensing Pulse Diagnosis Instrument for Chinese and Western Medicine,” 2014.
[13] Chung, Yu-Feng, et al. "Possibility of quantifying TCM finger-reading sensations: II. An example of health standardization." European Journal of Integrative Medicine 4.3 (2012): e263-e270.
[14] Velik, Rosemarie. "An objective review of the technological developments for radial pulse diagnosis in Traditional Chinese Medicine." European Journal of Integrative Medicine 7.4 (2015): 321-331.
[15] J.W. Ye, “Baseline Elimination by Wavelet and Cubic-Spline Interpolation for 3DPM and Improvement of L-Cube Pulse Mapping Recognition,” 2016.
[16] Zong, W., et al. "An open-source algorithm to detect onset of arterial blood pressure pulses." Computers in Cardiology, 2003. IEEE, 2003.