phyonyeinchan/quantum-cardiology

Quantum Machine Learning for Causal Inference in Cardiovascular Epidemiology

This repository applies a Variational Quantum Classifier (VQC) using IBM Qiskit to estimate propensity scores for causal inference, utilizing clinical records of heart failure patients. This methodology provides a quantum-enhanced approach to risk stratification and confounding adjustment in observational epidemiological data.

Live Interactive Research Dashboard

👉 CLICK HERE TO VIEW THE FULL INTERACTIVE REPORT

🎖 Verified IBM Quantum Credentials

The underlying core competencies driving this research are backed by professional IBM Quantum certifications:

<table> <tr> <td align="center"> <a href="https://www.credly.com/badges/c13b4d21-4c38-4b28-9128-90fa3984c81d/publicurl"> <img src="https://images.credly.com/size/220x220/images/6474ba18-3b83-4a34-878d-158e5869a20d/Quantum20Machine20Learning.png" width="110" alt="Quantum Machine Learning"><br/> _{<b>Quantum Machine Learning</b>} </a> </td> <td align="center"> <a href="https://www.credly.com/badges/1b2f698d-ac5a-46d1-9b55-64f295028d33/publicurl"> <img src="https://images.credly.com/size/220x220/images/3e026f05-3d63-44a0-9535-ce5c9089a268/Fundamentals20of20Quantum20Algorithms.png" width="110" alt="Fundamentals of Quantum Algorithms"><br/> _{<b>Quantum Algorithms</b>} </a> </td> <td align="center"> <a href="https://www.credly.com/badges/79bc0d5d-81cd-4aaf-8dc8-cb3d28820d70/publicurl"> <img src="https://images.credly.com/size/220x220/images/60cbe993-f35f-4b98-b7f6-8cd51233fe2a/image.png" width="110" alt="Basics of Quantum Information"><br/> _{<b>Quantum Info Basics</b>} </a> </td> <td align="center"> <a href="https://www.credly.com/badges/dd8fea00-87ff-41a0-bd52-cacbc74b9238/publicurl"> <img src="https://images.credly.com/size/220x220/images/e934a013-956f-421a-9a3d-668da9eac7ae/General20Formulation20of20Quantum20Information.png" width="110" alt="General Formulation of Quantum Information"><br/> _{<b>General Formulation</b>} </a> </td> </tr> <tr> <td align="center"> <a href="https://www.credly.com/badges/8164796f-e343-45ac-a66e-e57eafbae640/publicurl"> <img src="https://images.credly.com/size/220x220/images/f3ae0407-0eaa-4af7-ab15-0735c81af889/Foundations20of20Quantum20Error20Correction20.png" width="110" alt="Foundations of Quantum Error Correction"><br/> _{<b>Error Correction</b>} </a> </td> <td align="center"> <a href="https://www.credly.com/badges/bc31a16e-8469-4fc1-87d6-c8f2b61cf3b8/publicurl"> <img src="https://images.credly.com/size/220x220/images/53629652-f0f2-4bc8-abfd-444c366d3cf6/image.png" width="110" alt="Practical Introduction to Quantum-Safe Cryptography"><br/> _{<b>Quantum-Safe Crypto</b>} </a> </td> <td align="center"> <a href="https://www.credly.com/badges/7aa93544-ca9d-41fb-b00a-ede85fe567b5/public_url"> <img src="https://images.credly.com/size/220x220/images/f786ecd2-dce6-49d1-b7c6-6dcd832f8b14/image.png" width="110" alt="Quantum Business Foundations"><br/> _{<b>Business Foundations</b>} </a> </td> <td></td> </tr> </table>

🌟 Research Highlights

• —Epidemiological Context: Focuses on adjusting confounding biases in observational cardiovascular cohort studies.
• —Quantum Advantage Exploration: Utilizes high-dimensional Hilbert spaces via Quantum Feature Maps to capture complex, non-linear interactions among clinical covariates.
• —Hardware Optimized: Developed and executed natively using local quantum simulators optimized for Apple Silicon (M4 architecture).

🛠 Project Structure

• —preprocessing.ipynb: Data ingestion, cohort inspection, train/test splitting, and clinical feature scaling using StandardScaler.
• —quantum_causal_model.ipynb: End-to-end implementation including 10-Qubit Quantum Circuit design, embedding clinical parameters, and VQC model training.

📊 Technical Specifications

• —Dataset: Heart Failure Clinical Records (299 patients, 10 clinical confounding features).
• —Quantum Mapping: 10-Qubit configuration leveraging ZZFeatureMap (linear entanglement, 1 rep).
• —Ansatz & Optimization: RealAmplitudes (1 rep) driven by the COBYLA classical optimizer (max 20 iterations).

🚀 How to Run Locally

1. Prerequisites

Ensure you have Python 3.9+ and the required packages installed on your local environment:

pip install qiskit qiskit-machine-learning pandas scikit-learn numpy

2. Execution

Open the project folder in VS Code and execute the notebooks using your local Jupyter kernel:

# Execute end-to-end training
# Current deployment optimized for local Apple Silicon GPU/CPU simulator execution