Related Experiment Video
Updated: Jul 20, 2025
10:28
Dynamic Digital Biomarkers of Motor and Cognitive Function in Parkinson's Disease
Published on: July 24, 2019
15.2K
Externally validated deep learning model to identify prodromal Parkinson's disease from electrocardiogram
Ibrahim Karabayir1, Fatma Gunturkun2, Liam Butler1
1Cardiovascular Section, Department of Internal Medicine, Wake Forest School of Medicine, Medical Center Boulevard, Winston-Salem, NC, 27157, USA.
Scientific Reports
|July 29, 2023
Summary
Artificial intelligence can now predict Parkinson's disease (PD) risk years before diagnosis using standard electrocardiograms (ECGs). This AI model identifies early PD markers, enabling timely interventions and clinical trial enrollment.
Area of Science:
- Neurology
- Cardiology
- Artificial Intelligence
Background:
- Parkinson's disease (PD) prodromal stage lacks established electrocardiogram (ECG) markers.
- Early detection of PD is crucial for timely intervention and management.
Purpose of the Study:
- To develop a generalizable, fully automatic artificial intelligence (AI) model for predicting prodromal Parkinson's disease risk using ECGs.
- To assess the model's predictive accuracy up to five years before clinical diagnosis.
Main Methods:
- A case-control study utilized standard 10-second, 12-lead ECG data from two independent cohorts (MLH and LUC).
- A novel one-dimensional convolutional neural network (1D-CNN) was developed and validated for PD risk prediction.
- The AI model was compared against feature engineering-based models, with subgroup analyses for sex, race, and age.
Main Results:
- The 1D-CNN model achieved an external validation AUC of 0.74 for predicting PD within 6 months to 1 year prior to diagnosis.
- Predictive accuracy improved closer to the diagnosis time window (AUC 0.69 for 6 months-3 years, AUC 0.67 for 6 months-5 years).
- The AI model using raw ECG data outperformed traditional feature engineering approaches.
Conclusions:
- Standard ECGs analyzed by AI can effectively identify individuals in the prodromal stage of Parkinson's disease.
- This approach offers a cost-effective method for population-level early detection and facilitates enrollment in disease-modifying therapeutic trials.
