CMOS標準製程之振動式自感測微懸臂梁應用於人體血漿中凝血時間的監測

Monitoring blood clotting time in human plasma by an CMOS oscillated piezoresistive microcantilever sensor

石一弘(Yi-Hong Shih)

指導教授 ： 張培仁 ； 顏毅廣 ； 黃榮山

國立臺灣大學/工學院/應用力學研究所/碩士(2017年)

https://doi.org/10.6342/NTU201703869

摘要

本研究使用CMOS商業化標準製程技術開發出自感測壓阻式微懸臂梁凝血感測器，搭配快速傅立葉轉換演算法分析訊號，應用於抗凝血劑用藥監測的評估方式─凝血酶原時間(PT)與部分活化凝血酶原時間(aPTT)的監測。現代人生活型態轉變及飲食不當是引起心血管疾病的重要因素，為了預防及治療由血液與血管壁異常作用造成的血管栓塞，改變日常飲食習慣外，病患亦須依賴抗凝血劑的治療。然而抗凝血劑若用藥不當，將造成出血等副作用，故患者需定時監測血液的狀態是否在正常範圍內。目前因為血液的檢驗屬醫療等級，患者需到醫療單位接受生醫檢測，然而從檢體的處理、運送、儀器排程到最後取得報告的時間冗長，在臨床上難以達到因應患者血液狀態，即時調整用藥劑量的效果。若能夠配合病患作息進行即時監測的定點照護技術會是一大貢獻，如何將醫療等級的血液檢測發展到能在臨床或病患家中監測將是未來重點。
量測方式參考分析血液凝固狀況的Sonoclot分析儀，使用激振器以固定振幅、固定頻率來驅動壓阻式微懸臂梁，使其在待測樣品中振動，利用血凝樣品Reynolds number變化時微懸臂梁的受力情形也會產生變化為基礎，擷取感測器之訊號來推知待測樣品性質的變化。搭配快速傅立葉轉換演算法得知特定頻率的振幅值，本研究以此值來反映出微懸臂梁的受力情形，進而分析並得到凝血酶原時間和部分活化凝血酶原時間。
利用不同濃度的甘油水溶液進行實驗來了解微懸臂梁在不同Reynolds number環境中的受力情形，實驗結果得知微懸臂梁阻值改變量的10 Hz振幅與Reynolds number有正相關的趨勢，得到∆R/R_0 =0.7366[1/Re]+4.2643方程式來表示本微懸臂梁在液體中振動的表現且線性度相當良好(R2=0.9651)，證實振動式微懸臂梁感測器能夠分辨不同Reynolds number之液體且準確度相當高。接著將量測標的改為實際凝血情形，利用自行設計之演算法來處理訊號後可得知在凝血反應過程中特定頻率的振幅變化情形，以振幅明顯驟升所需時間做為微懸臂梁所量測到之凝血酶原時間與部分活化凝血酶原時間，三重複實驗結果得到第一級血凝品管液的Son-PT為9.83秒(標準差為0.85秒)、Son-aPTT為32.17秒(標準差為3.27秒)；第二級血凝品管液的SonPT為24秒(標準差1.22秒)、Son-aPTT為47.83秒(標準差為3.12秒)；第三級血凝品管液的Son-PT為37.83秒(標準差3.4秒)、Son-aPTT為71.5秒(標準差為2.55秒)。此量測結果與目前醫療院所使用之量測儀器比較結果，在95%信賴區間內無法證明兩者之間的差異性，證實使用微懸臂梁量測凝血時間具有相當高的準確性；同時量測結果具有特定的圖形，可監測凝血反應時纖維蛋白的生成與纖維蛋白聚集形成血塊的階段搭配微懸臂梁能夠感測不同Reynolds number的實驗結果，本感測器能夠監測凝血反應過程的黏度變化情形，且使用快速傅立葉演算法能夠讓我們有效的去除雜訊得到適當的資訊。
本研究開發之振動壓阻式微懸臂梁感測器屬於半導體技術，故有可微型化與成本低的潛力，且後端訊號處理也可設計於晶片中，在定點照護領域中有很大的發展空間。量測凝血反應的技術中能夠量化描述凝血過程的並不多，如Sonoclot分析儀，但目前除了凝血時間的量測外，其他血液資訊的準確度屬研究階段。綜觀以上，本研究利用血液Reynolds number的變化來描繪凝血反應過程，具有很大的發展潛力。

關鍵字

微懸臂梁 ； 壓阻 ； 凝血 ； 雷諾數

並列摘要

This study has developed a real-time coagulation monitoring sensor by using an CMOS externally vibrated, self-sensing piezoresistive microcantilever for disposable point-of-car coagulation device. With the increasing use of oral anti-coagulant drugs and increasing adverse drug events, the need for point-of-care coagulation devices has become necessary. Prothrombin time (PT) and activated partial thromboplastin time (aPTT) are measure of blood coagulation, and both of them are index for anticoagulant therapy to determine the blood condition in coagulation reaction.
In this study, the measurement was performed by vibrating the piezoresistive microcatilever immersed in the sample liquid at a fixed frequency of 10 Hz and fixed amplitude of 145 μm. The acquired signal of resistance change in microcantilever was processed by Fast Fourier Transform algorithm, and the resistance amplitude in 10 Hz indicated the amount of force exerting to the cantilever. In coagulation reaction, the viscosity of samples was sharply changed due to the clot formation, and the increased force can be sensed when the resistance amplitude in 10 Hz rises. Prothrombin time and activated partial thromboplastin time can be obtained by the time needed for fibrin clot formation. The method was initiated by Sonoclot analysis.
The amplitude of resistance in the specific frequency was found in a good linear correlation with absolute viscosity changes of glycerol/water solutions (R2 > 0.94). It was also found that the amplitude-k absolute viscosity curve behave differently in very low kinematic viscosity, probably due to the decrease in viscous drag of l absolute viscosity fluids. Also, the Reynolds number correlation can be achieved to present the relation of vibrated microcantilevers in sample liquid. Thus, ∆R/R_0 =0.7366[1/Re]+4.2643 (R2 = 0.9651) was derived to successfully describe the relation between acquired signals and vibrated Reynolds number. In addition, three types of commercially standard human plasma samples for measurement of coagulation prothrombin time were used for characterizing microcantilever sensors. The measured results of resistance amplitude in specific frequency with specific patterns of signature indicated the viscoelastic changes in blood coagulation reaction process. In coagulation reaction of human plasma control level 1, the PT measured by the microcantilevers was 9.83 sec with std. of 0.85 sec, the aPTT was 32.17 sec with std. of 3.27 sec; PT = 24 sec with std. of 1.22 sec, the aPTT was 47.83 sec with std. of 3.12 sec in human plasma control level 2; and PT = 37.83 sec with std. of 3.4 sec, the aPTT was 71.5sec with std. of 2.55 sec in human plasma control level 3. Compare with commercial coagulation device, the PT and the aPTT showed an excellent agreement between the microcantilever sensor and commercial device in 95% confident range. All results lay in the ranges of references. The experiment results demonstrated that the PT and the aPTT can be measured by vibrated microcantilevers accurately and precisely. Thus, this microcantilever sensor has demonstrated the real-time measurement for point-of-care coagulation monitoring, and shown its potential in miniaturization for personal diagnosis.

並列關鍵字

microcantilever ； piezoresistance ； clotting time ； Reynolds number

參考文獻

45. 林豪駸, 利用快速傅立葉轉換系統分析自感測壓阻式微懸臂梁於凝血反應之監測. 臺灣大學應用力學研究所學位論文, 2015: p. 1-86.

1. Basu , D., et al., A Prospective Study of the Value of Monitoring Heparin Treatment with the Activated Partial Thromboplastin Time, in New England Journal of Medicine. 1972. p. 324-327.

2. Investigators, B.A.A.T.f.A.F., The effect of low-dose warfarin on the risk of stroke in patients with nonrheumatic atrial fibrillation. N Engl j Med, 1990. 323(22): p. 1505-1511.

4. Heneghan, C., et al., Self-monitoring of oral anticoagulation: a systematic review and meta-analysis. The Lancet, 2006. 367(9508): p. 404-411.

6. Plesch, W., et al., Results of the performance verification of the CoaguChek XS system. Thrombosis research, 2008. 123(2): p. 381-389.