以體外透析瘻管實驗模型來探究血管通路狹窄程度與血流聲音之關係

Study on the Relationship between the Stenosis Degree and Blood Flow Sound of the Vascular Access Using an Experimental Model of Extracorporeal Vascular Access

殷嘉駿(Chia-Chun Yin)

指導教授 ： 王家鍾

義守大學/醫學院/生物醫學工程學系/碩士(2021年)

摘要

透析瘻管對於洗腎病人是他的第二生命，一旦透析瘻管出現狹窄時，會引起栓塞，造成血管不通，甚至完全阻塞，而須重建手術另造一條瘻管。目前評估動靜脈瘻管是否狹窄，初步是由臨床醫護人員藉由觸診、聽診加以檢查。目前全球洗腎人口不斷攀升的趨勢下，若能以簡易的瘻管血流聲音來判斷血管通路窄化程度，是最可行及最急迫的研究課題。本論文建立了一體外透析模擬系統，其中的人工瘻管是利用3D列印製成，正常管徑為7.2 mm，阻塞管徑分別7.0mm、6.0mm、5.0mm、4.0mm、3.0 mm，並利用電子聽診器在不同血流、血壓、心率、口徑下，測得A點(血液流入阻塞血管前)、B點(阻塞血管中間)及C點(血液流出阻塞血管後)之正常口徑936筆及阻塞口徑4680筆10秒鐘血流聲音訊號(PAG)。於時域分析中，我們計算各筆血流訊號的平均振幅、平均能量及跨零點數目；於頻域分析中，將血流聲音訊號進行快速傅立葉轉換(FFT)，將血流聲音頻段分為 0 ~ 10、10 ~20、20 ~ 30、30 ~ 50、50 ~ 80、80 ~ 120、120 ~ 200、200 ~ 300、300 ~ 500、500 ~ 4k Hz，並計算各頻段之能量百分比。由時域分析結果顯示，相較於正常口徑下，我們發現阻塞口徑所測得的血流聲音之振幅大小、平均能量、及跨零點數目均會隨著管徑的縮小而變小或變少，例如正常口徑(7.2 mm)之血流聲音平均振幅顯著大於及阻塞口徑(5.0 mm)( 0.13±0.03 vs 0.10±0.02 V, p<0.001)。於頻域分析中，我們發現由正常口徑A點所測得的血流聲音訊號在低頻段(0~10 Hz)的能量百分比顯著大於阻塞口徑(35.14±5.40 vs 25.85±4.16V %，p<0.001)；B點所測得的血流聲音訊號相較於正常口徑下低頻段訊號的表現些微下降，中高頻段(30~80Hz)訊號上升，依管徑的不同呈顯著差異(5.82±1.37 vs 6.88±1.99，p=0.03)。C點點所測得的血流聲音訊號在低頻與中高頻段有明顯表現，在高頻段(120~200 Hz)有部分管徑有些微的表現(3.87±1.11 vs 3.02±1.01，p=0.005)。研究得知，隨著瘻管從正常管徑至阻塞程度越來越嚴重，血流聲音訊號頻段分佈會從低頻段轉至中高頻段。因此，利用體外透析模擬系統透過改變管徑的分析結果，可見間接作為動靜脈瘻管是否有無阻塞現象之評估，以利病人在瘻管尚未完全阻塞時，先至心血管中心做處置，以延長瘻管壽命。

關鍵字

動靜脈瘻管 ； 血流聲音 ； 體外透析模擬系統 ； 時域分析 ； 頻域分析

並列摘要

Dialysis fistula is their second life for dialysis patients. Once the dialysis fistula becomes narrow, it will cause embolism and finally partially or completely block the fistula or nearby blood vessels. So, reconstructive surgery is required to create another fistula. At present, the patency of the arteriovenous fistula is initially checked by clinical medical staff through palpation and auscultation. With increasing tendency of the global population with dialysis, it is necessary to use an available technique to observe the narrowing degree of vascular access by means of the photoangiogram (PAG) signals.
In this thesis, an extracorporeal dialysis simulation system was established. The artificial fistula tube with different diameters were constructed using the 3D printing method. The normal tube diameter was 7.2 mm and the obstruction tube diameters were 7.0, 6.0, 5.0, 4.0 and 3.0 mm, respectively. Under different conditions with several levels of blood flow, blood pressure, heart rate and caliber, both normal (936 records of 10 seconds) and stenotic (4680 records of 10 seconds) PAG signals at point A (before the blood flow entering the blocked vessel), point B (the middle of the blocked vessel) and point C (after the blood flow leaving the blocked vessel) were acquired using an electronic stethoscope. In the time domain analysis, we calculated the average amplitude, average energy, and cross-zero number of each 10-sec PAG signal. In the frequency domain analysis, the frequency spectrum of the 10-sec PAG signal was divided into 10 bands: 0 ~ 10, 10 ~ 20, 20 ~ 30, 30 ~ 50, 50 ~ 80, 80 ~ 120, 120 ~ 200, 200 ~ 300, 300 ~ 500 and 500 ~ 4K Hz. The time-domain analysis results showed that compared with the normal caliber, the average amplitude and average energy of the PAG signal measured from the obstructed caliber of 5.0 mm both were smaller than the normal-size fistula (0.13±0.03 vs 0.10±0.02 Volt, p<0.001; 35.14±5.40 vs 25.85±4.16V %, p<0.001). In the frequency domain analysis, we found that the PAG signal measured at point B of the normal-aperture fistula had significantly greater energy percentage at low frequency band of 0~10 Hz than the blocked aperture (21.45±3.40 vs 40.05 ±7.53 %, p<0.001), while at the medium-to-high frequency bands the PAG signal from the normal-size fistula showed smaller energy percentage. We found that certain amount of energy at the low frequencies in the PAG signals measured from the stenotic fistula was shifted to the medium-high frequencies. In summary, the proposed extracorporeal dialysis simulation system can be applied to assess the patency of the fistula by means of analyzing the PAG signals measured at different positions. The finding will indirectly help to diagnose the patency conditions of the vascular access in clinical settings and possibly to extend the usage life of the dialysis fistula.

並列關鍵字

Arteriovenous Fistula ； Photoangiogram ； Extracorporeal Dialysis Simulation System ； Time-Domain Analysis ； Frequency-Domain Analysis

參考文獻

[1] 美國UNITED STATES RENAL DATA SYSTEM年報，HTTPS://WWW.USRDS.ORG/, 2017.

[2] 杜翌群(南臺科技大學電機工程系暨研究所)，「 洗腎瘻管照護輔助裝置之研製)」， 科技部專題研究計畫成果 報告(MOST 105-2221-E-218-008, 2016/08/01~2017/09/30)。

[3] 陳維聆(高雄榮民總醫院)，「 在模擬動脈系統的動靜脈瘻管模型以音頻技術探討血管通路窄化程度變化之功 能性方法」， 國科會專題研究計畫成果報告(MOST 105-2218-E-075B-001, 2016/03/01~2017/02/28)。

[4] 林康平(中原大學電機工程學系)，「以白光 LED 顯微造影技術在裸鼠腫瘤微循環血管之血流速分析研究」，99 年度國科會專題研究計畫(NSC 99-2221-E-033-012-MY3, 2010/08/01~2013/07/31)。

[5] Chen WL, Kan CD, Lin CH, Mai YC. ‘‘Generalized Regression Estimator Improved the Accuracy Rate of Estimated Dialysis Accesses Stenotic Condition on In-Vitro Arteriovenous Graft Experimental Model. IEEE Access.’’ Digital Object Identifier 10.1109/ACCESS.2018.2802479, 2018.