چاپ صفحه |  صفحه نخست سامانه |  نویسندگان |  اطلاعات تفضیلی |  دانلود مقاله | دانشگاه علوم پزشکی تبریز |
| نویسنده | نفر چندم مقاله |
|---|---|
| گلاویژ خدری زاده | اول |
| رضا علی اکبری خویی | سوم |
| رویا دولت خواه | پنجم |
| توحید جعفری کشکی | چهارم |
| سعید موسوی | ششم |
| عنوان | متن |
|---|---|
| کلمات کلیدی | Survival Analysis, Censored time, Fuzzy logic |
| خلاصه مقاله | It is often the case that the estimation of covariates' effect on survival time in highly censored data, has a large standard error and is prone to bias. This study proposes a method based on Fuzzy Product Limit Estimator (FPLE) to address these problems. The FPLE is used to construct membership functions for the event and censored times. Then the area under the membership functions is calculated as estimations of event times. Using the estimated event times, a linear model is fitted to the constructed data. Simulations are performed to study the precision of the proposed method and compare the method with Cox and Exponential regression. Finally, the method is implemented on survival data of colorectal cancer (CRC). Simulated survival data are generated from a Cox model with two covariates and censoring rates of 30, 50, and 70%. Then the data is clustered based on the covariates. The proposed method is applied to the clustered and non-clustered data. The bias and standard error of covariates' effect estimates with Cox, Exponential and Proposed methods were compared. The proposed method with clustering has the lowest bias and standard error. It is suggested to use the proposed method for handling survival data with heavy censoring |
| نام فایل | تاریخ درج فایل | اندازه فایل | دانلود |
|---|---|---|---|
| Enhanced_Certificate.png | 1404/10/09 | 7761716 | دانلود |
| FPLE.pdf | 1404/10/09 | 5163193 | دانلود |
