📝 Project Description This project focuses on automated liver tumor segmentation from CT medical images using Deep Learning. It leverages the TransUNet model, a hybrid of CNN and Transformer architectures, to achieve accurate pixel-level segmentation of tumor regions.
🔑 Key Features Data Preprocessing: Normalization, resizing, and noise reduction of CT scans.
Feature Extraction: Deep learning methods to capture hierarchical tumor patterns.
Segmentation: Precise delineation of tumor boundaries.
Stage Identification: Automated tumor size calculation and clinical stage classification.
Web Application: User-friendly interface for uploading scans and viewing results (segmentation mask, tumor size, stage).
🎯 Goal To assist healthcare professionals by reducing manual effort, minimizing diagnostic errors, and providing timely insights for treatment planning.