We're developing a deep learning model to detect cancer cells in lymph nodes. Typically, lymph node cancers are diagnosed through histological examination after biopsy. To automate the diagnosis of cancer cells, we'll utilize deep learning, specifically the transfer learning technique. This approach will enable earlier detection of cancers and save lives. Currently, the available dataset for training and testing is limited. Additionally, analyzing histopathological images requires significant computational power and performance. Furthermore, in the traditional analysis method, pathologists need to perform staining to identify cancerous areas clearly, which incurs additional costs. We have conducted extensive research to address these major challenges in diagnosing cancer cells. Moreover, we're utilizing explainable AI techniques to provide clear explanations for our predictions regarding the cancer-affected segments. So our project aims to evaluate cancer cell predictions using our model, which is designed to be cost-effective and requires low computational power and performance. Additionally, we ensure explainability by visualizing the cancer segments in the provided histopathological images. Ultimately, we plan to develop a website where histopathologists can easily access and utilize our model for their diagnostic purposes.
- E/17/230 - S.Nishankar - e17230@eng.pdn.ac.lk
- E/17/256 - S.Piriyaraj - e17256@eng.pdn.ac.lk
- E/17/358 - N.Varnaraj - e17358@eng.pdn.ac.lk
- Dr. Pratheepan Yogarajah - p.yogarajah@ulster.ac.uk
- Dr. Swarnalatha Radhakrishnan - swarnar@eng.pdn.ac.lk