A Study of Feature Extraction with Back Propagation Neural Network and Support Vector Machine in Acute Appendicitis

• Main Authors: Kuo-Hung Yang, 楊國宏
• Other Authors: Tung-Hsu Hou
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/52980058108551678513

碩士 === 國立雲林科技大學 === 工業工程與管理系 === 102 === Appendicitis is the most common cause of acute surgical abdomen in Emergency Room (ER).In the early, the acute appendicitis are similar with peritonitis, genital diseases of female, gastrointestinal disease and urinary disease. According to the literature review the percentage of negative appendectomies has been reported to vary in 25% to50 %. The important factors that influence the acute appendicitis were applied based on the index of symptoms, signs and laboratory data were applied in feature extraction with principal component analysis. Also, back propagation neural network and support vector machines combined with principal component analysis were used to establish the prediction model for acute appendicitis. 188 appendectomy cases form regional hospital located in southern Taiwan from were used to test to proposed system by four-fold cross validation. The result indicates that the proposed back propagation neural network combined with integrated type of principal component analysis is the best method to predict acute appendicitis, and the accuracy reaches 95.74%. This computer aided clinical evaluation system can help the hospital to decrease the numbers unnecessary operation of the derivative cost and dissension, assist the doctor to balance the rate of ruptured or negative appendectomies and the uncertainty of operation, and reduce the anxiety for the patient and their family.