Major project Healthcare & Hospitals Case Studio

AI ECG screening for routine clinics

Inspired by publicly reported work at Mayo Clinic

Detect several heart conditions from 12-lead ECGs and design how a doctor would act on an alert.

The brief

Mayo Clinic has described ECG-AI that flags heart conditions from a routine ECG, and a trial where doctors got an alert in the health record suggesting a follow-up test. Build a multi-label ECG model on a large public 12-lead dataset, calibrate it, check it across age and sex, and design the alert a doctor would see. Good looks like: AUROC per condition on the official test folds, calibration curves, subgroup results, and an alert design with a clear follow-up step and limits.

Read the case first

What Mayo Clinic has said in public about this kind of work, on its own pages:

Make it your own

Design the screening for an Indian setting: a district hospital, a health camp or a company health check, where cardiologists are few. Decide who sees the alert, what they do next, and what happens if there is no echo machine nearby.

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

Load the 12-lead ECGs and diagnostic labels
Use the recommended train/validation/test folds
Train a 1D CNN or ResNet multi-label model
Calibrate and choose thresholds per condition
Report subgroup results by age and sex
Design the clinician alert and write a model card

Free, open data to build it with

It takes a minute with your email. AiTeky Pass (₹999/month or ₹9,999/year) is free for your first 30 days.