Predicting future heart failure with health records and wearables
Official title: Prediction of Heart Failure Onset using Multimodal Data Analysis, Deep Learning and Commercial Wearables
Funded at University of Michigan at Ann Arbor · NIH-11166497 · Data updated 2026-11-12
This project uses medical records, clinic ECGs, and smartwatch heart-rate data to try to predict who may develop heart failure within the next year.
Quick facts
| Grant type | Career grant |
|---|---|
| Study type | NIH-funded research |
| Funding institution | University of Michigan at Ann Arbor NIH-funded |
| Lab location | 1 site (Ann Arbor, United States) |
| Project start | September 1, 2021 |
| Project end | August 31, 2026 |
| Funding year | 2025 |
| Award (this year) | $162,853 |
| NIH project number | 5K01HL155404-05 |
| Project ID | NIH-11166497 on NIH RePORTER |
Who could benefit from this research
✓Could help
Ideal candidates are adults receiving care at Michigan Medicine who do not yet have diagnosed heart failure but have risk factors or concerning heart information in their records.
✕Less likely to help
People who already have a diagnosed and active heart failure condition are unlikely to benefit from a prediction-focused project.
What this research studies
You would share your medical record and, if selected for the wearable part, wear a smartwatch for seven days so researchers can collect heart rate and ECG-like data. The team will use deep learning to combine your health record information with clinic ECG and wearable heart-rate variability signals to look for patterns that appear before heart failure begins. They will train and test the models using past patient data from Michigan Medicine and then compare whether consumer wearables can replace clinic measurements. If the wearable data works well, it could make early warning tools easier to use outside clinics.
Why it matters
Potential benefit: If successful, this could identify people at high risk of heart failure earlier so preventive care and lifestyle changes can start sooner.
How similar studies have gone: Other studies using EHRs or wearables have shown promise for detecting heart-rhythm issues, but combining EHR, clinic ECG/HRV, and consumer wearable data to predict heart failure a year ahead is relatively new.
Where this research is happening
-
University of Michigan at Ann Arbor
Ann Arbor, United States Active
Researchers
- Principal investigator: Ansari, Sardar — University of Michigan at Ann Arbor
- Study coordinator: Ansari, Sardar
Related research
Other NIH-funded projects on the same topics.
-
Personalized CRT planning using advanced heart imaging and AI
A scientist is testing a new AI-assisted imaging approach to help doctors choose the best CRT device (BIVP or LBBAP) for heart failure patients with specific conduction problems.
-
Using AI to Understand and Improve Treatment for a Type of Heart Failure
This project uses advanced computer methods to better understand different forms of heart failure with preserved ejection fraction, aiming to find the most effective treatments for each person.
-
Using deep learning to improve heart failure care quality and reduce disparities.
This study is looking at how to improve care for people with heart failure, especially for Black individuals who often face more challenges, by using advanced technology to better understand their health records and find ways to provide better support and treatment.
-
Predicting when heart failure might occur using data from wearables and health records
This study is looking to help people at risk of heart failure by using smart technology and health records to predict when it might happen, giving doctors a chance to step in and help up to a year in advance.
-
Using deep learning to improve lung ultrasound assessments for heart failure in emergency patients
This study is looking at how to make it easier and faster for doctors in the Emergency Department to check for heart failure in patients by using smart technology to analyze lung ultrasound images, helping you get the care you need more quickly.
-
AI-supported heart failure diagnosis and management in Rwanda
Aiming to help people with suspected heart failure in Rwanda get faster, more accurate diagnoses and better care using AI-guided ultrasound and care protocols.
About this research
- This is an active NIH-funded research project — typically early-stage science, not a clinical trial accepting patient enrollment.
- Some NIH-funded labs run parallel clinical studies or seek volunteers for related work. To check, contact the principal investigator or institution listed above.
- For full project details, budget, and progress reports, visit the official NIH RePORTER page.