Study Rationale:
The nervous system regulates heart rate in response to situations such as stress and emergencies. Dysfunction of part of the nervous system controlling the heart can cause changes in electrocardiogram (EKG), which measures the electrical activity of the heartbeat, in people with Parkinson’s disease (PD). In this study we will look for evidence of EKG changes in the earliest stages of PD, even before other classic signs of disease are observable.
Hypothesis:
We hypothesize that artificial intelligence (AI) applied to EKGs can help identify people in very early stages of PD, before they show other classic signs of PD.
Study Design:
We will use EKG data from the Honolulu Asia Aging Study cohort to build AI-based models that can identify people at the earliest phase of PD. We will then see how well these models work on a different set of patients with PD from Methodist LeBonheur Healthcare System in Memphis, Tennessee.
Impact on Diagnosis/Treatment of Parkinson’s disease:
If successful, our project will yield a noninvasive, cheap and easy-to-use tool for identification of people likely to develop PD. Such a tool could identify participants earlier for clinical trials, at a time when drugs are more likely to be effective.
Next Steps for Development:
Next steps following this proposal will be to test our AI-based model in additional populations in order to fully understand its ability to predict PD. If successful, we hope to make this model available to researchers for ongoing and future clinical trials.