On Friday 22 February, Alexander Kuleshov met with Russia’s head cardiologist, Sergei Boitsev, and reached an agreement on the transfer of one hundred thousand patient data entries, the result of a long-term pilot project and years of work.
As part of this project, the National Medical Research Center of Cardiology (NMRCC) will use a database containing approximately one hundred thousand text entries from the period of 2011 to 2018, thirty thousand numerical electrocardiogram entries from 2014-2018, and three thousand MRI images from 2014-2018.
This pilot project between Skoltech and the NMRCC has developed an accurate scale to predict the risks of cardio-embolic stroke based upon clinical summaries from the patients’ medical records. The mathematical model takes into account a number of factors: sex, age, arterial hypertension, type-2 diabetes, heart failure, serious atherosclerosis, as well as early thromboembolic complications.
A team of doctors and mathematicians is focusing on finding a means to predict cardio-embolic stroke after the development of atrial fibrillation (or even without development of atrial fibrillation). This includes the creation of an algorithm for screening and categorizing at-risk patients, for whom prescribing anticoagulants is advisable to prevent cardio-embolic stroke before verifying atrial fibrillation.
Using artificial intelligence, this project aims to map patients’ medical histories – demographic, clinical, lab – as criteria for future research and for testing potential predictors of atrial fibrillation and ischemic stroke. The goal is to process MRI images of the heart based on machine learning, but with expert engagement as well patients’ ECG records. Automated analysis will also cover MRI imaging and ECG records to identify potential predictors of atrial fibrillation.
+7 (495) 280 14 81