Classification and Precision Diagnosis of Endometrial Cancer - Survey

Abstract

Diagnosis of Cancer type is as important as the right classification of benign and malignant tumor cells. It is also very critical to achieve precision diagnosis to help doctor’s find better clinical outcomes. Endometrial type of a cancer that starts in   the uterus and found at an early stage as it frequently produces abnormal vaginal bleeding and hence it becomes critical to not only classify the tumor but also to detect the type and grade of the cancer.

In the United States of America, endometrial cancer is a top five leading cancer with fifty-two thousand unique cases reported in the year 2014. This number raised to around sixty thousand in the year of 2016 and these numbers are further estimated to increase to around sixty-two thousand in 2019.

The dataset used is composed of radiology and pathology images from Corpus Endometrial Carcinoma (CPTAC) Patients. This paper uses the Machine learning approaches that will classify the cell as malignant or benign. Deep learning approaches are used to provide Precision diagnosis that provides better outcomes to the clinicians. Convolutional Neural Network algorithms have been successfully used in diagnosis of Brain tumor, Skin Cancer and similar approaches can be used to tackle this problem.