Cardio Vision, a Ukrainian startup, has developed a smart service to combat heart disease. It all started with the fact that the Australian owner of a network of hospitals turned to Ukrainian students, and their cooperation turned into a successful project.
How it all started
The owner of a network of hospitals in Australia turned to students of the Ukrainian Catholic University (UCU) with a problem: patients had to wait a long time for the diagnosis of cardiac data – there were so many of them that the specialists simply did not have time to process them.
Bogdan Petrishak and Marichka Dobko (2019 grads) worked on artificial intelligence (AI) in the medical industry, starting with their first year at UCU. They were joined by an UCU teacher, Oles Dobosevich, (Deputy Dean of the Faculty of Applied Sciences of the Ukrainian Catholic University) who had great expertise in the development of artificial intelligence. Together, they found a solution.
As a result, they created the Cardio Vision project, which works on the basis of artificial intelligence. By the way, in 2019, the team received its first serious victory – 1st place in the international AI for Good Challenge 2019 competition from Microsoft among developers, students, and specialists in data who are looking for innovative solutions based on AI.
Now the team is developing clinical tests for a launch in Australia, and soon in Ukraine.
How Cardio Vision Works
Every third death in the world is caused by cardiovascular disease. These are heart attacks, strokes, cardiac arrhythmias, and heart failure. But most often people die as a result of coronary heart disease (a problem with blood supplied to the heart muscle).
Several different tools are used to diagnose coronary insufficiency. One of the most common is CT coronary angiography.
Cardio Vision is a web service that allows doctors to download CT images of a patient’s coronary angiography and, using artificial neural networks, determine the degree of blockage of the coronary arteries, the so-called level of stenosis (circulatory disorders). The service also helps identify patients in need of emergency medical care in the first place and informs the doctor about this.
What is unique
The coronary angiography CT procedure generates a very large number of images (1000-1500 for one patient). The hospital, serving 10-20 patients, generates about 20 thousand images daily. This is a very large amount of data that doctors simply do not have time to physically analyze. They spend about half of their work time just interpreting these pictures.
The service is software available online, and special equipment is not required.
The uniqueness of Cardio Vision technology lies in the fact that the system automatically determines the level of stenosis in CT images of coronary angiography. This is not done by any of the many competitors in the market.
How it works
Artificial intelligence plays a key role in Cardio Vision’s technology solution. The doctor uploads the patient’s heart images to a web service. The algorithms analyze all the data and give predictions of the level of stenosis in each clinically important section of the coronary tree: a healthy section, an insignificant level of stenosis, or a dangerous stage of capping. Thus, the specialist receives predictions of the general condition of the coronary artery and can instantly determine the stage of the disease of a particular patient.
Also, radiologists have the opportunity to open each image separately and highlight all the areas that the neural network pays attention to in the final forecast. Basically, highlighted areas are blockages of various types that cause narrowing of the coronary arteries. This option is very important, because users of the service will be able to receive an interpretation of algorithm predictions in this way.
Cardio Vision does not make decisions for the doctor. This is not a product that will take away the work of specialists or replace them. This is a doctor’s decision support technology that will make routine work easier, faster and simpler.
When hospitals will be able to use it
Cardio Vision is still in the testing phase. The team is developing it and conducting clinical trials. According to plan, during the first half of 2020 the project will be able to successfully pass all stages of testing and enter the Australian market first, and then the Ukrainian one.
Join us for UCU Days – Sep 8-14 and you can learn about UCU achievements from first hand. Thank you for your support!