A framework for the classification of cervical cancer stages from mri using customized cnn and transfer learning

Abstract

In the early stages of cervical cancer, treatment is possible, and the patient can return to a normal lifestyle. It is crucial to obtain accurate cervical cancer staging results in order to determine the best course of therapy. Automatic disease detection and diagnosis applications are more necessary for diagnosing the cervical cancer stage. CNNs play a more significant role in object detection and classification than other neural networks. This is because CNNs performance is on par with radiologist skills in some cases. The main objective of this thesis is to build a framework for the automatic classification of cervical cancer stages from MRI. MRI plays a major role in staging cancer disease in the medical field. It requires a knowledgeable radiologist to accurately diagnose the stage of cervical cancer. The proposed framework based on deep learning CNN assists radiologists in diagnosing the stage of cervical cancer disease. There are many ways to classify an image or determine if an organ is diseased. In this research, three CNN based models are used for the classification of stages of cervical cancer from MRI. The final step in medical image analysis is to analyze the results and determine how they can be used. Automatic classification helps classify cervical stage from the MRI. It is considered a fundamental operation for planning treatment strategies. In the first part, Transfer Learning using VGG16 is used in the staging of cervical cancer MRI. This step helps to classify the cervical cancer stages from MRI without constructing a deep learning model from scratch. newline newline newline