Our Motivation

What inspired us!

The overall incidence of spontaneous ICH worldwide is 24.6 per 100,000 person-years. Approximately half of the mortality occurs within the first 24 hours, highlighting the critical importance of early and effective treatment of the same.

The deep learning frameworks have enabled more accurate and faster detection. The radiologists could scale their diagnostic efforts with the increased CPU and GPU processing power available. Time is a matter of grave concern for medical diagnosis, and early identification of the location and type of any hemorrhage can potentially add years to the life of a patient.

Next

Our Architecture

Innovation in the model

Model is taking average predictions from ResNet-50 with SVM and XGBoost.

Next

Future Scope

Discussion

These applications will have the flair to impact the diagnostic approaches of clinicians and health care systems and the ability for individuals to lay eyes on changes to their health in real-time.

Next

Hang On!

Few description

Code

Github repository

Documentation

Paper work

Novelty in Model

ResNet-50 with SVM & XGBoost