碩士 / 雲林科技大學 / 工業工程與管理研究所碩士班 / 97 / Predictive models for clinical results in elderly hip fracture surgery:a compassion study of two methods between logistic regression and artificial neural network.
Student:Chen-Chiang,Lin Advisors:Yung-Ching,Liu
Institute of Industrial Engineering and Management
National Yunlin University of Science & Technology
ABSTRACT
This study was to investigate the recovery status after operation one year later in elderly hip fracture patients. There were three models of this study to predict mortality rate, hip joint recovery status and general health recovery status. Model 1 was the predictive model of mortality rate. The charts of 286 patients were completely reviewed and the collected datas were analyzed with logistic regression and artificial neural network. to build up a predictive model of mortality rate. Model 2 was the predictive model of hip joint recovery status. The charts of 71 patients were completely reviewed. The 71 patients pilled in a eldly hip fracture questionnaire and Harris hip score questionnaire. The collected datas were analyzed with logistic regression and artificial neural network to build up a predictive model of hip joint recovery status. Model 3 was the predictive model of general health recovery status. The charts of 71 patients were completely reviewed. The 71 patients pilled in a elderly hip fracture questionnaire and Short Form-36 questionnaire. The collected datas were analyzed with logistic regression and artificial neural network to build up a predictive model of general health recovery status. The results showed that the predictive accuracy of artificial neural network was higher than the predictive accuracy of logistic regression. The applications of this study can provide patients, family and doctors an objectively predictive data. to increase the accuracy of decision making and to decrease the legal problems.
Keywords:Elderly hip fracture. Logistic regression, artificial neural network.
Identifer | oai:union.ndltd.org:TW/097YUNT5030027 |
Date | January 2009 |
Creators | Chen-Chaing Lin, 林鎮江 |
Contributors | Yung-Ching Liu, 柳永青 |
Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
Language | zh-TW |
Detected Language | English |
Type | 學位論文 ; thesis |
Format | 69 |
Page generated in 0.0014 seconds