Challenge background
Early this year Omdena’s Morocco team used a computer vision model to interpret ECGs, specifically acute ST-Elevation MI. However, the team struggled to find enough labelled data and subject matter experts. By providing specialist cardiology input, this collaboration between Morocco and London will expand the training data set and optimise the model.
The problem
The project’s objective is to use ECG data to build a computer vision system that will be able to learn from cardio specialist doctors and perform classification and/or risk scoring. This solution could be used to manage CVD risk in rural LMICs where there are no specialized cardiac services. However, when it comes to deployment we must be mindful of access to the internet.
Goal of the project
- Collect Knowledge about ECG interpretation.
- Get a dataset of ECG with classification.
- Train a computer vision model to be able to classify the ECG.
- Deploy the model in a real environment with specialist supervision to test the accuracy.
Project timeline
- 1
Week 1
Collect data and knowledge about access to ECGs in LMICs
- 2
Week 2
Optimise CV Model and develop a risk score
- 3
Week 3
Test with cardiology specialists and get feedback
- 4
Week 4
Deploy the App in Cloud Application Platform
What you'll learn
The participants will be able to apply AI for one of the most recurrent scenarios in healthcare: classification of medical imaging. E-health is starting to deploy AI and this project will bring valuable learning opportunities.