利用逆運算疊代法評估頸動脈疾病患者之臨床數據預測其血管狹窄程度

本研究是將頸動脈狹窄之病患的臨床資料運用逆運算疊帶法的運算分析方式來預測臨床病人頸動脈狹窄的程度。首先定義五個臨床上頸動脈狹窄的重要危險因子，年齡(Age)、低密度膽固醇(LDL-C)、平均動脈壓(MAP)、飯前血糖(Sugar ac)、C反應蛋白(CRP)，再整合成一個十六項的一階非線性方程式作為預測的基準。經由STATISTICA 7.0軟體進行逆運算疊帶法演算。將217位頸動脈疾病患者之臨床資料正規化，亦即將所有數據歸化至-1~+1之間，再用上述演算法進行預測，即可算出頸動脈狹窄程度的預測數值。結果為最終損失函數值Φ=2.354，決定係數R2=0.9352，變異數為87.46%。最後，再以55例具有頸動脈疾病的病患，進行臨床驗證預測，結果表現出高度的吻合，R2=0.875。而危險因子中CRP因子貢獻值最大為0.7204，因此為最主要因子，反之，因子年齡的貢獻最小為0.0256。所以結果顯示以原始頸動脈疾病患者之臨床危險因子數據所建成的資料庫，再經由具有相同疾病患者進行驗證，發現實際和預測狹窄程度具有高度的吻合度。未來臨床上可以運用逆運算法，以非侵入性生理數據事先預測患者頸動脈狹窄程度，得以提早治療，減少缺血性腦中風發生之危險，並且可以減少醫療上不必要的浪費。關鍵字: 頸動脈狹窄、逆運算疊帶法、C反應蛋白(CRP)、正規化、腦缺血性腦中風

關鍵字

頸動脈狹窄；逆運算疊帶法； C反應蛋白(CRP) ；正規化；腦缺血性腦中風

並列摘要

The clinical data of patients with carotid stenosis syndrome were analyzed using the revised inverse problem to predict the degree of stenosis in this study. Six factors defined as Age, Low-Density Cholesterol (LDL-C), Mean Arterial Pressure (MAP), Sugar AC before feast and C-Reactive Protein (CRP) were adopted as high risk parameter to develop a first-order nonlinear semi-empirical formula with 16 items according to a revised inverse problem run by STATISTICA 7.0 software. Accordingly, six of high risk clinical data of 217 patients with carotid artery syndrome were collected and normalized to a range within -1~+1, respectively. This was essential to ensure that the theoretical calculation derived from each high risk factor was judged and evaluated under same criteria and no intrinsic bias was accompanied in the prediction. The results showed a high convincible prediction with final loss function value Φ 2.354, determination coefficient R2 0.935, and variance 87.46%. Furthermore, another group of 55 patients with same carotid artery syndrome were recruited and verified to ascertain the accuracy, and the obtained results showed a high degree of agreement, R2 0.875 as well. The major contribution of the prediction came from the CRP, since its coefficient was 0.7204 in the 16 items semi-empirical formula, thus CRP is the most dominant of the high risk factor among all six factors, whereas Age is the most minor factor for its smallest coefficient, 0.0256 as well. The theoretical prediction was accomplished according to the six high risk factors and provided helpful information for authorities to create appropriate policy for concerning the issue of public health care. Keywords：Carotid artery stenosis, Revised inverse problem, C-Reactive protein (CRP), Normalize, Brain ischemic stroke.

並列關鍵字

Carotid artery stenosis ； Revised inverse problem ； C-Reactive protein (CRP) ； Normalize ； Brain ischemic stroke

參考文獻

1. 劉濟弘、李宗海，林口長庚醫院醫學中心-認識頸動脈狹窄，長庚醫訊2015;第39卷9期:18-19頁。

Google Scholar

2. Ralph LS, Greg A, Karen F, et al. Guidelines for Prevention of Stroke in Patients With Ischemic Stroke or Transient Ischemic Attack. Circulation 2006; 113 : e409-e449.