ECG heartbeat classifier
Inspired by publicly reported work at Mayo Clinic
Classify single heartbeats from ECG recordings into normal and abnormal types.
The brief
Mayo Clinic has written about AI that reads ECGs to spot heart disease early. Start with a basic building block: classify single heartbeats from a public ECG database into normal and abnormal types, deal with the imbalance, and show which mistakes matter most clinically. Good looks like: per-class recall and a confusion matrix, a patient-wise split so the same patient is not in train and test, and a note on why this is not a diagnostic tool.
Read the case first
What Mayo Clinic has said in public about this kind of work, on its own pages:
- Spotlight on early detection of 3 heart diseases using ECG-AI (Mayo Clinic News Network) ↗
- Trial demonstrates early AI-guided detection of heart disease in routine practice (Mayo Clinic News Network) ↗
Make it your own
Design it for where you are: a low-cost single-lead device in a rural clinic, with noisy signals. Add realistic noise to the test data and show how the model holds up.
AITeky has no link with Mayo Clinic, which has not reviewed or endorsed this brief. The brief only draws on what the company has published, uses open data rather than its data, and is a learning exercise: don't present your project as Mayo Clinic's work or system.
What you will build
- Explore heartbeat signals and classes
- Split by patient, not by beat
- Train a 1D CNN and a feature-based baseline
- Report per-class recall and a confusion matrix
- Write a note on clinical limits
Free, open data to build it with
Sign up free to get the data and links
It takes a minute with your email. AiTeky Pass (₹999/month or ₹9,999/year) is free for your first 30 days.