應用於行動照護之動態取樣全數位心電訊號擷取電路

Abstract

近年來慢性疾病逐年上升，預防醫學顯得越來越重要。為了可以全天候的監控人體的生理訊號，並減低人們往來醫院的不便，利用可攜式電子進行全天候的行動照護成為一個很好的選擇。此研究設計一個應用於行動照護心電訊號擷取電路之動態取樣全數位類比數位轉換器。以壓控振盪器為主的類比數位轉換器(VCO-based ADC)可達到高的面積使用效率，以及高的數位整合度，這些都是行動醫療照護所需要的。此外，隨著製成的微小化，量化於時域的VCO-based ADC擁有較好的解析度。此研究設計一有效位數高於8位數的VCO-based ADC並以90奈米製程實現。此ADC可支援單一及多個通道的心電訊號擷取電路。 以單一通道應用而言，取樣頻率為1k赫茲，此ADC可達到10.48位數的有效位數此時的功率消耗為6.08瓦特。而在多通道的應用下，此設計可操作在10k赫茲底下，以提供8個通道的心電訊號擷取。可在維持有效位數為9.64位數的情況下，降低單一通道的功率消耗至0.62瓦特。 為了更進一步地降低ADC的功率消耗，我們根據生理訊號的特性提出動態取樣的機制。利用偵測訊號的變化程度，來調整ADC的解析度，使得在變化程度低的區間ADC的功率消耗可大為減少。此機制可在失真率小於5%的情況下達到53% 的功率降低量。與現有的其他機制相比，此方法與有最高的功率降低效率。

Heart attack becomes the top cause of death in the U.S. in 2011. The preventive medicine of the heart disease becomes more and more important. To reduce the effort for people going to the hospital, the mobile healthcare device is applied to monitor the electrocardiogram (ECG) signal. Here we design a VCO-based ADC for the ECG acquisition circuit in mobile healthcare applications. The VCO-based ADC has high area efficiency and high integration ability with the digital circuit which are suitable for the mobile device. And also benefits from the technology scaling. Using 90-nm CMOS process technology, the ENOB of more than 8-bit VCO-based ADC is implemented. It can support both single channel and multi-channel ECG acquisition circuit application. For the single channel, the ENOB of ADC can reach 10.48 bits in sampling frequency of 1k Hz. The power consumption is 6.08 uW. For the multi-channel application, our design can operate in sampling frequency of 10k Hz and be turned off while sampling is finished. In the simulation result, the power of VCO-based ADC is 0.62uW/ channel for 8 channels application and achieves 9.64 bits of effective number of bit. To reduce the power consumption of our ADC, we apply the dynamic sampling technique to sense the variation of signal. Then we adjust the resolution of ADC based on the sensing result. This technique can reduce 53% power of ADC while keeping distortion less than 5%. Compare with other technique this work has best power reduction performance under the same distortion factor.