It always challenging for medical students and new doctors to interpret electrocardiograms (ECGs) given how often they are used in clinical practice.
What it does
HeartBeat helps them to interpret ECGs and differentiate between cardiac arrhythmias. It is a tool that they can use on top of their existing clinical knowledge in order to confirm the interpretation of the ECG.
How we built it
We used a cardiac arrhythmia dataset from UC Irvine and taught our classifier to differentiate between normal ECGs and ECGs showing cardiac arrhythmias (eg heart attacks, heart blocks, sinus tachycardia, etc.).
Challenges we ran into
It was challenging to figure out what to teach the classifier, how to input the data, and how to structure it neatly. Connecting everyone's work was challenging too (front-end, machine learning algorithm, and back-end).
Accomplishments that we're proud of
We're proud that we could work together to get our project done within 24 hours! Teaching a machine to distinguish between the presence and absence of cardiac arrhythmias and classifying it was technically challenging.
What we learned
We learnt a lot about Python, ECGs, scikit-learn, and machine learning!
What's next for HeartBeat
Instead of uploading csv files to input the data, it would be cool to just upload images so that pictures are just needed in order to interpret the ECGs.