主動式脈搏量測暨重現系統

Abstract

在中醫的診斷中，把脈為診斷疾病的重要依據。然而中醫師為病患把脈時，無法將患者的脈搏跳動觸感轉換成量化的數據儲存，也難以得知該患者前次的脈象情況。目前大多數的脈搏量測裝置僅針對單一按壓力度下的脈搏量測，使用者難以任意調整按壓力度進行量測。在脈搏重現裝置方面，多數的裝置本體體積龐大且難以攜帶，而且多以進行固定的脈搏波形呈現。 本研究開發一套脈搏量測暨重現系統，量測及重現裝置皆基於體積小且方便攜帶的條件下開發。量測裝置方面，使用者能夠施加不同按壓力度，並且方便調整位置進行脈搏量測所開發的脈搏量測裝置，利用壓力感測器量測按壓力度，同時用薄膜式振動感測器(polyvinylidene difluoride, PVDF)量測人體脈搏原始振動訊號。使用田口法針對脈搏量測裝置進行最佳化設計。在重現裝置方面，依照使用者施加的按壓力度下，能夠重現由脈搏量測裝置所測得的脈搏訊號。 本研究使用類神經網路(artificial neural network, ANN)建立重現裝置電壓輸入訊號與脈搏重現訊號的轉換關係。結果顯示，浮、中、沉取之脈搏重現訊號，與原始脈搏訊號的判定係數R2分別為0.966、0.973、0.968。由此可證明本研究所開發之三指脈搏重現裝置，能夠正確重現人體脈搏振動訊號。

In the traditional Chinese medicine (TCM) diagnosis, pulse is an important indicator for medical diagnosis. However, when TCM doctors are taking the pulses for a patient, it is hard for them to quantize the pulse data. Without the quantitative pulse data, it is hard to obtain the past pulse history. Current measurement devices mostly focus on developing single pressure pulse measurement, making it less flexible to measure the pulse. As for the current reproduction devices, the size of the devices is large, which makes them unhandy and immovable. Since the portability is the key element in device development, this research develops a small-size pulse measurement and reproduction system. When using the measurement device, users are able to apply multiple pressures and adjust the position of the device according to their needs. In the measurement device, the pressure sensor is used to measure the pressure force, and the piezo film vibration sensor is used to measure the original vibration signal of human pulse. This research also applied Taguchi method for the measurement device optimization. The pulse signals obtained from the measuring device are further reproduced by the proposed reproducing device. This research uses an artificial neural network to establish the conversion relationship between the voltage input signal of the reproducing device and the pulse reproduce signal. The experiment results show that the three-finger pulse reproducing device developed by this research can correctly reproduce the human pulse vibration signal accurately.