A STUDY ON ENDOMETRIAL CARCINOMA USING NOVEL ASSAYS

Abstract

The present study a data mining technique, which will be useful to narrow down the data without losing the factors, an algorithm was developed to enumerate the factors which seem to be responsible for endometrial carcinoma to enumerate the high risk areas, which can be assessed by any health manager to counteract endometrial carcinoma. By applying the logistic regression model, it is inferred that the factors responsible for endometrial cancer are menopause, parity, socio-economical status and education out of which two of these factors are responsible globally whereas the other two factors are responsible locally in nature, where the health managers should take proper precautions to counteract the occurrence of endometrial carcinoma in the region. After applying this method it was found that age, diet, parity, marital status and menopause are responsible for the occurrence of endometrial carcinoma. This study also employed image processing techniques which are Basedon box counting algorithm and 2 D discreet wavelet; which is to differentiate normal and abnormal region of interest by the radiologist. Both methods are complimentary to each other and are supporting aid to the radiologist which is also useful for oncologist to take the proper decision. These methods avoid biopsy which is a non-invasive and will not affect the cognization of the patient. In addition to this the methods are used to early diagnosis and to track the disease once it occurred. Finally the treatment of the patient can be evaluated using these methods to trace the efficacy of the drug. From the study it is concluded that the tumor has heterogenecity and also leads to the heterogenecity in genomics subtypes; variations in the expression of growth factors and free anti-angiogenic factors and variations in the tumoral micro environment.