The company aims to help improve arrhythmia detection through donation of tremendous dataset gifted to the general public through PhysioNet.
Icentia released its largest public dataset to date - a record-breaking 11,000 patients’ ECG records and 2 billion beats recorded over one week . The CardioSTAT recorder was developed by Icentia, a company dedicated to improving arrhythmia detection using ECG data.
Their new dataset includes annotations of each beat type by certified technologists, such as normal beats, premature atrial contraction, and premature ventricular contraction, and rhythms such as normal sinus rhythm, atrial fibrillation, and atrial flutter.
The prevalence of arrhythmia are becoming alarmingly common, with 2.7 to 6.1 million patients in the U.S having Afib alone, with estimates of this number rising to a staggering 12.1 million by 2030. Despite new automated methods of detecting arrhythmia using machine learning, ECG technologists and cardiologists still encounter ECG anomalies that do not fit the known arrhythmia morphology, hence the need for human expertise and intervention to compliment computer-based learning.
To assist medical professionals, continue to investigate these rhythm anomalies and have a dataset of raw ECG signals to compare and match them to, Icentia recorded 2 billion labeled beats from 11k patients. The data is completely anonymized, extracted from patients who wore a CardioSTAT recorder for one week on average, ensuring each patient recorded a significant and statistically sizeable number of beats and rhythms for maximum data authenticity and trustworthiness.
To perform this ground-breaking data gathering project, Icentia utilized its’ CardioSTAT recorder, a compact, single-use and portable ECG recorder engineered to continuously monitor long term electrocardiograms.
Once the raw data was extracted from the patients CardioSTAT recorder, Icentia's certified technologists performed annotation using their proprietary analysis tools. They thoroughly analyzed the complete record and labeled the beat and rhythm types. This final data was then reviewed and approved by a senior technologist before becoming a publicly available dataset on PhysioNet.
Through the public availability of this data, Icentia hopes to help cardiologists and ECG specialists fight this public health challenge of arrhythmia, through early diagnosis and treatments.
To know more about CardioSTAT, visit www.cardiostat.com or to download the 11k dataset, visit the Physionet repository at www.physionet.org