肥胖病人膽結石生成之預測因子研究

The study in Prediction of Gallbladder Disease among Obese Patients

背景：類神經網路是一個非常普遍的技術，可以用來處理相當複雜模式的資料。目前並沒有應用類神經網路在發展肥胖病人膽結石發生的風險值預測研究，而肥胖是膽囊疾病及膽結石的危險因子之一。因此本研究試圖利用羅吉斯迴歸以及類神經網路來回溯性分析台灣地區肥胖病人膽囊疾病之危險因子。方法：本研究收集1999年到2005年到本院施行肥胖手術共117位病患，利用倒傳遞類神經網路建構預測膽結石發生的因子。30個輸入變數包括了人口統計值、實驗室資料以及膽結石的組織病理學結果。結果：類神經網路比起羅吉斯迴歸呈現了較佳的預測準確率以及較低的第二型誤差。透過資料採礦的技術找出的危險因子包括：舒張壓，慢性發炎和異常的醣類代謝及膽固醇沈著變性。結論：類神經網路是一項有用的工具，透過實驗室資料和病理學特徵等變數可以用來預測膽囊疾病發生以及在肥胖病人膽結石生成的風險因子。類神經網路的預測績效是優於傳統的統計方法。

Background:Obesity is a risk factor for gallbladder disease and gallstone. The authors retrospectively analyze the prevalence and risk factors of gallbladder disease using logistic regression and artificial neural networks among obese patients in Taiwan. Methods: Artificial neural network (ANN) is a very popular technique, which can detect complex patterns within data. They have not been applied to risk of gallstone development in obese population. We studied the risk factors associated with gallstones in 117 obese patients who were undergoing bariatric surgery between February 1999 and October 2005. ANN, constructed with three-layered back-propagation algorithm, were trained to predict the risk of gallstone development. Thirty input variables including clinical data (gender, age, BMI and associated diseases), laboratory evaluation and histopathologic findings of gallbladder were obtained from the patient records. The result was compared with a logistic regression model developed from the same database. Results: ANN demonstrated better average classification rate and lower Type II errors than those of logistic regression. The risk factors from both data mining techniques were diastolic blood pressure, inflammatory condition, abnormal glucose metabolism and cholesterolosis. The biological significance of inflammatory condition in obese population requires further investigation. Conclusion: ANN might be a useful tool to predict the risk factors and prevalence of gallbladder disease and gallstone development in obese patients on the basis of multiple variables related to laboratory and pathological features. The performance of ANN was better than traditional modeling techniques.