GO FROM BIG DATA
TO SMART DATA
HOW IT WORKS
Qardian implements deep neural networks in our software product, HEARO: Heart Evaluation for Algorithmic Risk-reduction and Optimization.
The algorithm works by using mathematical models to learn patterns in existing data, and then applying itself to predict outcomes for unfamiliar data.
To use the software, a doctor only needs to enter patient data metrics into our UI. Our software can be integrated with an EHR system, maximizing ease of use. With Qardian in an EHR, doctors will be able to view and analyze heart disease risk as part of a patient medical record.
Both HEARO-11 and HEARO-14 achieve 99% accuracy on test data. HEARO-11 is evaluated with the University of California Irvine machine learning repository dataset. HEARO-14 is evaluated with UT student athlete data gathered by Head Team Physician Dr. Chris Klenck.
HEARO's extremely high accuracy is suggestive of its potential to provide doctors with a vital piece in the complex medical puzzle of heart disease diagnosis.
Our method outperforms previously published results, and can provide a significant improvement upon current heart disease risk assessment methods.
HEARO benefits doctors by using deep learning to detect data relationships and output accurate results in under 5 minutes.
This research is described in further detail in the scientific paper, "On Deep Neural Networks for Detecting Heart Disease."