Bio-Inspired Optimization for Medical Data Mining

4Tumor Detection and Classification

Hermehar P.S. Bedi*, Sukhpreet Kaur and Saumya Rajvanshi

Department of Computer Science and Engineering, Chandigarh Engineering College - CGC, Mohali, Punjab, India

Abstract

Uncontrolled growth of abnormal tissues in the body is referred to as the tumor. Broadly, tumors can be classified into two types: benign and malignant. The World Health Organization (WHO) claims that in the year 2020, cancer alone contributed 10 million deaths, making it the leading death causing disease. National Cancer Registry Programme of Indian Council of Medical Research (ICMR), estimated cases of cancer in India have seen a rise from 1,392,179 in 2020 to 14,61,427 in 2022 and reported a mortality of 7,70,230 and 8,08,558 in 2020 and 2022, respectively, which is a significant increase. Till date, more than 120 types of brain tumors have been found, which have different shapes and sizes. Due to these factors, it is important to detect the tumors immediately. Various mechanical methods exist for the detection of tumors, such as magnetic resonance imaging (MRI), computed tomography (CT) scan, and positron emission tomography (PET) scan. All these scans send the waves into the patient’s body to obtain the images of the organ or body part affected. Images obtained from these scans are of Digital Imaging and Communications in Medicine (DICOM) format. Different techniques available for classification of the tumor, such as aegmentation, feature extraction, machine learning ...