利用血流動力學及自律神經檢查提早發現迷走神經性昏厥

Early Detection of Vasovagal Syncope in Tilt-up Test with Hemodynamic and Autonomic Study

陳俊安(Chun-An Cheng)

指導教授 ： 邱泓文

臺北醫學大學/醫學科技學院/醫學資訊研究所/碩士(2011年)

https://doi.org/10.6831/TMU.2011.00138

摘要

對於昏厥之病患需先詳細詢問病史是否有誘發因子，且排除心肌梗塞或心律不整，並且需與眩暈、癲癇及短暫性腻缺血區別，並安排傾斜床檢查以確定診斷。傾斜床檢查是評估反覆性原因不明昏厥之最佳工具。傾斜床檢查流程頇耗時長達45分鐘，利用延長傾斜床檢查之時間以增加診斷率，亦可使用血管擴張藥物以增加陽性率及縮短時間。但使用藥物檢查會使其專一性下降，增加假陽性機會。
因為迷走神經性昏厥病患之複雜性、多因子、異質性，故利用非藥物傾斜床檢查評估迷走神經性昏厥及解釋其生理病理發現。經由陽性病患及受詴正常者的血流動力學及自律神經參數了解迷走神經昏厥之生理變化分析其差異以了解其致病機轉。結果發現，迷走神經性昏厥病患的體重、身體質量係數、帄躺之帄均血壓較低；傾斜測詴下昏厥前期時，心跳、心率變異低頻/高頻比率上升、感壓反射敏感度下降較明顯、周邊血管阻力係數較低。於昏厥前其心跳加快、心率變異低頻/高頻比率上升代表交感神經活性增強，而較低的周邊血管阻力係數代表後負荷(血管壁交感神經活性)較低，昏厥時心跳減慢代表副交感神經活性增強。
並利用基本資料、身體組成、帄躺及傾斜早期血流動力學及自律神經參數，運用邏輯式回歸及類神經網路分析建立一模組早期預測其發生，發展一種早期偵測會在檢查時發生昏厥之病患，以免發生嚴重之意外。利用身體質量指數、帄躺時心跳、傾斜3分鐘心輸出係數、傾斜3分鐘左心做功係數、傾斜3分鐘帄均血壓訓練之類神經網路，訓練組敏感度為92%；專一度為95%，正確率為93.75%；驗證組正確率為100%。ROC闕值為0.517，ROC曲線下面積為0.979，整體敏感度為93%；專一度為96%，正確率為95.5%，模組狀況佳可提早辨識發生的病患。

關鍵字

迷走神經性昏厥 ； 邏輯式迴歸 ； 類神經網路

並列摘要

Syncope is a common disorder. The diagnosis of vasovagal syncope is according to history, heart rate and blood pressure change with postural challenge. The diagnostic test maybe use pharmacy to slow time of study has high sensitivity and false positive rate. We collected patients received tilt table test in Tri-service general hospital from 2005 to 2010, included total 1100 patients, 36 vasovagal syncope, excluded 2 older than 55 years-old, 4 without impedance cardiography. We selected 30 negative exam patients as control group.
Due to the complexity, heterogeneity, multiple factor, the pathophysological pathway was not fully know. The our compared study showed lower weight, body mass index, baseline mean blood pressure, rapid presyncopal heart rate, higher low frequency/high frequency ratio, lower baroreflex sensitivity and total peripheral resistance, then lower mean blood pressure and heart rate in syncopal period.
We used logistic regression and neural network to evaluate age, gender, height, weight, body mass index, and mean blood pressure, heart rate, heart rate variability ratio, blood pressure variability ratio, baroreflex sensitivity, stroke volume index, cardiac index, total peripheral resistance index, left ventricular work index, end-diastolic index, left ventricular ejection time and ejection time ratio during baseline and first 3 minutes,5 minutes during tilt table test to early detect vasovagal syncope with tilt table test. The model of ANN with body mass index, baseline heart rate, tilt up 3 minutes cardiac index, left ventricular work index and mean blood pressure can early predict VVS outcome. The Training group sensitivity is 92%; specificity is 95% accuracy is 93.75%; validation group sensitivity is 100%, ROC area under the curve of 0.979, the overall sensitivity is 93%; specificity is 96%, accuracy is 95.5%,the module performance is excellent can early detection of VVS patients.

並列關鍵字

vasovagal syncope ； logistic regression ； Artifical Neural Network

參考文獻

林麗玲，劉燕珠，王雲龍，林素娟，林國信(2010):捐血後不適就醫之探討。臺灣醫學，14，370-376。

陳惠芬(2008):迷走神經性昏厥與頭抬高傾斜床試驗。台南縣:長榮大學醫學研究所碩士論文。

Akselrod S, Gordon D,Madwed JB et al. Power Spectrum Analysis of Heart Rate Fluctuation: A Quantitative Probe of Beat to Beat Cardiovascular Control. Science. 1981;213:220-222.

Agelink WM, Malessa R, Bauman B, Majewski T, Akila F, Zeit T, Ziegler D. Standardized tests of heart rate variability: normal ranges obtained from 309 healthy humans, and effects of age, gender, and heart rate. Clin Auto Re.s 2001;11:99-108.

Beacher FD, Gray MA, Mathias CJ, Critchley HD Vulnerability to simple faints is predicted by regional differences in brain anatomy. Neuroimage. 2009;47:937-45.