Tumor Detection using Deep Learning in Organs Specific to Indian Predicament

Herein, a research study has been conducted to address the increasing prevalence of serious diseases, including uterine, breast, and colon cancer, particularly in Kerala, as identified through studies done by the state Government. The objective of this research work is to contribute to the urgent need for cancer detection, that would then help to reduce mortality rates and facilitate effective screening practices within the Indian context. To ensure a comprehensive analysis, a diverse dataset has been curated from multiple reputable web sources. The dataset consists of various types of medical imaging data, such as MRI scans, CT scans, and histopathological images, obtained from reliable repositories such as SipakMed, the National Cancer Institute, figshare, and the Cancer Imaging Archive (TCIA), among others. The dataset encompasses a wide range of cancer types and organ locations, including brain, breast, cervix, lung, colon, kidney, leukemia, and lymphoma. Each image in the dataset has been meticulously classified based on essential attributes, including tumor malignancy, location, and other relevant medical terminologies. In the research paper, several deep learning techniques, namely ResNet50, MobileNet, and VGG16, have been applied to perform tumor detection across multiple organs [3]. Through an extensive analysis, a comparative evaluation of the performance of each technique has been conducted, resulting in valuable insights for the effective utilization of deep learning in tumor detection across diverse organ types.