Skip to content

Scientific References & Theoretical Foundations

This document provides the bibliography, academic citations, and methodological foundations implemented in the AI ECG SDK (ecg_sdk).

1. Synthetic ECG Generation & Dynamical Cardiac Attractors

1.1 The McSharry Model

  • Citation: McSharry, P. E., Clifford, G. D., Tarassenko, L., & Smith, L. A. (2003). A dynamical model for generating synthetic electrocardiogram signals. IEEE Transactions on Biomedical Engineering, 50(3), 289–294.
  • DOI: 10.1109/TBME.2003.808805
  • Usage in SDK: Implemented in ecg_sdk.core.generator.generate_synthetic_ecg. Provides the circular limit-cycle attractor differential equations \(\dot{x}, \dot{y}, \dot{z}\) and the baseline Gaussian parameters for Lead II / MIT-BIH channel MLII.

1.2 Multichannel Vectorcardiographic Projections & Dynamical Extensions

  • Citation: Sameni, R., Clifford, G. D., Jutten, C., & Shamsollahi, M. B. (2007). Multichannel ECG and noise modeling: Application to maternal and fetal ECG signals. EURASIP Journal on Advances in Signal Processing, 2007, 1–14.
  • DOI: 10.1155/2007/43407

  • Usage in SDK: Formalizes the 3D cardiac dipole projection onto frontal (Einthoven/Goldberger) and horizontal (Wilson precordial) lead vectors.

  • Citation: Clifford, G. D., Shoeb, A., McSharry, P. E., & Janz, B. A. (2005). Model-based filtering, compression and classification of the ECG. International Journal of Bioelectromagnetism, 7(1), 158–161.

  • Direct Resource: ijbem.org/volume7/number1/158-161.pdf

  • Citation: Sayadi, O., Shamsollahi, M. B., & Clifford, G. D. (2010). Synthetic ECG generation and Bayesian filtering using a Gaussian wave-based dynamical model. Physiological Measurement, 31(10), 1309–1329.

  • DOI: 10.1088/0967-3334/31/10/002

2. Clinical Electrocardiography & 12-Lead Voltage Standards

  • Citation: Goldberger, A. L., Goldberger, Z. D., & Shvilkin, A. (2017). Goldberger's Clinical Electrocardiography: A Simplified Approach (9th Edition). Elsevier Health Sciences. ISBN: 978-0323401692.
  • Usage in SDK:
    • Frontal Plane Norms: Einthoven's Law (\(\text{III} = \text{II} - \text{I}\)), Goldberger augmented unipolar equations (\(\text{aVR}, \text{aVL}, \text{aVF}\)), and inverted \(a\text{VR}\) morphology (negative P, deep QS, inverted T).
    • Horizontal Plane Norms: Anatomical precordial \(R\)-wave progression from \(V_1\) (\(rS\) pattern) through \(V_3\) (\(RS\) transition zone) to \(V_5/V_6\) (\(qR\) lateral pattern).
    • Calibration Standards: Standard \(25\text{ mm/s}\) paper speed, \(10\text{ mm/mV}\) gain, and \(1\text{ mV} \times 0.2\text{ s}\) calibration pulse implemented in ecg_sdk.visualization.plotter.

3. Benchmark Datasets & Noise Models

4. Inter-Patient Evaluation & Explainable AI (XAI)

  • Inter-Patient Partitioning (DS1 / DS2 Protocol):
  • Chazal, P. de, O'Dwyer, M., & Reilly, R. B. (2004). Automatic classification of heartbeats using ECG morphology and heartbeat interval features. IEEE Transactions on Biomedical Engineering, 51(7), 1196–1206.
  • DOI: 10.1109/TBME.2004.827359

  • Eliminates data leakage by strictly separating patient records between training (DS1) and independent testing (DS2).

  • Explainability (XAI) in Deep Learning ECG:

  • Manimaran, G., et al. (2025). Explainable deep learning based techniques for ECG-based heart disease classification: A systematic literature review and future direction. Computers in Biology and Medicine, 199, 111324.
  • DOI: 10.1016/j.compbiomed.2025.111324
  • Grounding for the Wave-Alignment Score (\(WAS\)) auditing electrophysiological fidelity of saliency maps against true anatomical P-QRS-T boundaries.