A Platform to Assist Pathologists in the Cancer Detection Process
Naren SR, Vanitha Veerasamy, R Natrajan · 2023
Cancer is a complex disease affecting millions worldwide, and early detection and accurate diagnosis are crucial for effective treatment. Biopsy tiles are an essential diagnostic tool in cancer detection and play a significant role in cancer pathology. This research paper proposes the development of a website for multiple cancer detection using biopsy tiles, which would aid pathologists in affirming their findings in cancer pathology. The website would use machine learning algorithms to analyze biopsy tile images and provide a preliminary diagnosis for multiple cancer types. This paper outlines the gradio interface development, which entails using pre-trained neural networks like InceptionV3, ResNet50, Alex Net, and VGG16. The interface also helps the pathologists ensure how the model learns a specific image and can modify it according to their own purpose. Having given access to the code to the pathologists, they can even modify or integrate new data into the model. The proposed interface has the potential to improve the accuracy and efficiency of cancer diagnosis, which could ultimately lead to better patient outcomes.