AI Deep Learning Techniques for Automating Cancer Detection and Classifying Pictures of Various Cancers

Cancer identification is a huge issue for scientists and clinical specialists because it is the primary cause of death worldwide. The development of more accurate and efficient treatments is necessary because, despite the importance of early diagnosis, conventional cancer detection techniques may involve intrusive procedures and time-consuming analysis..To get around these challenges, this work employs AI-based techniques for automated a diagnosis of cancer with a focus on deep learning models. The proposed fine-tuned Densenet121 model, VGG16, VGG19, ResNet34, Inception-V3, Resnet50, MobileNetv2, and other convolutional neural networks (CNNs) are evaluated on image files for seven different cancer kinds: cervical, lung and colon, lymphocytic leukemia (ALL), breast, kidney, brain, and oral. Images are initially subjected to segmentation processes, after which contour features are extracted and metrics such as area, epsilon, et perimeter are computed. The suggested model has the greatest validation precision (99%), the lowest loss (0.001), and RMSE criteria with the lowest values (0.031 for training as well as 0.042 for validation)), according to extensive testing of the models. The suggested model proved to be the most effective one in the study, suggesting that AI-based techniques could increase the accuracy of cancer detection.