Classification of Ovarian Tumor using Deep Learning Techniques

Cancer of the ovaries is one of the most challenging malignancies to diagnose accurately due to its varied tumor types and complex morphology. This study leverages advanced deep learning techniques, integrating CNN’s, GeoProteoNet & VGG models that enhanced with transfer learning to improve detection precision. Tumor is classified based on the shape of the tumors and the area of the origin of the tumor and some other parameters. There are majorly three types of tumors. They are Germ cell tumor, Stromal tumor and epithelial tumors. With the deep learning model we can classify and test the type of tumor and accurate tumor segmentation. The hybrid approaches address challenges like limited datasets and variability in tumor morphology by incorporating augmented data and multi-model genomic-proteomic analysis. The method achieves over 90% accuracy(99.7%), promising significant advances in the ovarian cancer diagnostics and reduced diagnostic delays.