Tech Stack: PyTorch, Python

Expert’s cases:

1. Trained segmentation model of Atlas of ASPECTS for MRI images using PyTorch Lightning. Performed hyperparameter tuning with Optuna. The dice score was 87%. Posted article on Medium

2. Implemented a service for processing MRI pictures and sending coordinates of segmented areas with OpenCV

3. The service assisted the doctors to segment and detect the deceased areas of the brain. Trained segmentation model of hemorrhagic stroke blood type for CT images using 2.5D technique. It helped to decrease the weight of the model on the server and still classify blood type and segment blood as good as with 3D images

4. Made a live dashboard with Streamlit. Used Apache Airflow for DAGs management. The dashboard helped to track the statistics of diagnoses by each week. Used pandas, matplotlib, seaborn, pylab. for EDA