Principal Component Logistic Regression in the early diagnosis of ovarian cysts Research and Software Development.doc

 PAGE \* MERGEFORMAT 5 Principal Component Logistic Regression in the early diagnosis of ovarian cysts Research and Software Development [Keywords:] Logistic regression; ovarian neoplastic cysts; non-neoplastic ovarian cysts; Abstract: Objective: Logistic regression analysis, principal component methods for early diagnosis of ovarian cysts, and this in order to facilitate the implementation of software development, out-patient diagnosis, and to improve the accuracy of the differential diagnosis of the disease and efficiency. Methods: Logistic regression model first collinearity diagnostics, and then filter out the differential diagnosis of ovarian cysts are statistically significant indicators for the inspection and the establishment of Logistic regression model, using the model of the early differential diagnosis of ovarian cysts. Use delphi 7.0 software program development. Results: The pregnancy times, parity, abortion exist between the number of moderate collinearity relations; among pregnant times, shape, wall structure, the remaining 11 indicators were included in model; retrospective discriminant line rate of 8786%, forward-looking difference coincidence rate of 8514%, discriminant effect is good. Conclusions: Logistic regression of principal component implements the early differential diagnosis of ovarian cysts and development of the software can be used for patient diagnosis. Keywords:: Logistic regression; ovarian neoplastic cysts; non-neoplastic ovarian cysts; delphi 7.0 software, the non-neoplastic ovarian cysts because the hypothalamic-pituitary-gonadal axis dysfunction in the formation of such factors as retention cysts [1], belongs to physiological tumor-like lesion, surgical treatment is not easy, but it is from the shape, size, etc. and pathological authenticity ovarian tumor neoplastic cystic tumor is very similar to the clinical are particularly vulnerable to confusion, secondary inspection (laparoscopy, CT, etc.) the effect is not satisf