zaina-ml/cardioiq
An AI Cardiovascular Risk Detection App - PRESIDENTIAL AI CHALLENGE
What's novel
An AI Cardiovascular Risk Detection App - PRESIDENTIAL AI CHALLENGE
Code Analysis
5 files read · 3 roundsA PyTorch-based system that uses a custom multi-scale ResNet with SE blocks to classify ECG beats from MIT-BIH data and fuses this with synthetic patient vitals to estimate cardiovascular risk.
Strengths
The code demonstrates strong substance with a well-architected, non-trivial ECG model featuring multi-scale convolutions and sophisticated temporal augmentations. The separation of concerns between ECGNet and CardioRiskNet is clean, and the data loading logic for handling time-series medical signals is robust.
Weaknesses
The project lacks an entry point (`app.py`) and configuration file, making it difficult to deploy or run without manual setup. Reliance on synthetic patient data for risk assessment limits real-world applicability, and there are no visible unit tests to verify correctness.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
10
Contributors
1
Files
62
Active weeks
6
Repository
Language
Python
Stars
0
Forks
0
License
MIT