Project Info:

I pioneered an automated brain tumor segmentation solution using AWS, implementing CNNs like U-Net and DeepLab on diverse medical imaging datasets. AWS services, including S3, EC2, Lambda, and SageMaker, enabled a robust data pipeline and scalable model deployment via Kubernetes. This project highlights my skills in medical image analysis, cloud computing, and DevOps, emphasizing my dedication to healthcare innovation in my portfolio.

Result Obtained :

Our U-Net model achieved an impressive Dice coefficient of approximately 0.85 during testing, indicating excellent agreement between predicted and ground truth tumor segmentations in MRI scans. We successfully integrated this tool with hospital systems while ensuring strict adherence to medical data privacy standards and regulations.

Project Details:

• Technologies:Pyhton, Tensorflow, pytorch, imgaug, AWS, Docker, Gith, mlflow, Kubernetes