����ɫ��

Heart muscle diseases, called cardiomyopathies, come in many different forms and often develop differently, so diagnosing a specific type can be challenging. Even in similarly appearing hearts, the cause of the disease may be different, and individuals my require different treatments to prevent serious complications like heart failure, life-threatening heart rhythms or death.

A team of researchers, led by the Cumming School of Medicine’s Drs. James White, MD, Alessandro Satriano, PhD, with Dr. Russ Greiner, PhD, of the , is tackling this issue head on with an innovative artificial intelligence (AI)-based diagnostic tool. The tool uses cardiac magnetic resonance imaging (CMRI) to diagnose the presence and type of cardiomyopathy.

The team recently received a $1.1 million Canadian Institutes of Health Research (CIHR) project grant to validate the effectiveness of their software with an international study. Called AID-MRI, involving 2,500 patients from sites in North America, South America, Europe and Asia, the study will examine if the software can learn to effectively diagnose diseases in people with different ethnicities across multiple countries.

The study, one of the largest to demonstrate the value of AI, will apply machine learning techniques to data captured by the Cardiovascular Imaging Registry of Calgary (CIROC), including more than 25,000 CMRI tests, to produce models that can diagnose different cardiomyopathies. 

“Over the last several years, there have been new treatments introduced to treat specific causes of cardiomyopathy, but many people aren’t getting these treatments because of the difficulty of diagnosis,” says White, a professor in the Department of Cardiac Sciences. 

How the software works

The software works by analyzing routinely captured CMRI of an individual’s beating heart and comparing its shape and patterns of movement to models built from others suffering from specific types of cardiomyopathies. The AI software considers thousands of data points to provide the most likely diagnosis to physicians within seconds. 

“Each heart condition has a unique pattern, or fingerprint, that we can train our AI-based software to recognize,” says White. “Ultimately, this can act as a virtual consultant to assist physicians to make the correct diagnosis when it matters most.” 

The unique data generated by these models can also be used to predict the future occurrence of heart complications, such as heart failure, atrial fibrillation and life-threatening arrhythmias. Researchers are planning to expand the software to identify those at highest risk of developing these complications.

The information can then be used to diagnose the disease and predict what may happen for each individual patient, allowing physicians to match the appropriate therapies to their patients.

“Our approach converts routinely available images generated by any MRI scanner into a standardized data model to assist physicians in making the correct diagnosis and treatment decisions for individual patients,” says White. “We are excited to confirm the accuracy of our techniques in an international setting.”

Atrial fibrillation in diabetes

Atrial fibrillation (AF) is the most common form of cardiac arrhythmia and can negatively affect a person’s quality of life. For some, it leads to heart failure and stroke.

Individuals with both Type 1 and Type 2 diabetes are 40-60 per cent more prone to developing AF, and at a higher risk of death and worsened quality of life compared to diabetics without AF.

AF in individuals with diabetes may develop for reasons that are different from AF in other conditions, and, although AF is similar in people with Type 1 and Type 2 diabetes, AF is not identical in these groups either. The mechanisms that underlie these differences are not well understood, which impedes the development of more personalized treatment options for AF in diabetics.

Dr. Robert Rose, PhD, a professor in the departments of �����Ի��� in the   (CSM), and members of his lab, are tackling this challenge.

Rose received a $994,500 CIHR Project Grant to continue studying the mechanisms and possible therapeutic approaches to atrial fibrillation in individuals with both Type 1 and Type diabetes. Type 1 diabetes is an autoimmune disease in which the immune system destroys the insulin-producing cells of the pancreas leaving people reliant on external insulin.

, the Rose lab demonstrated that mice with Type 1 diabetes are highly susceptible to developing AF and impaired electrical conduction in the heart due to loss of function of a critical ion channel in the atria called the sodium channel. The study went on to show that controlling insulin levels prevents these changes in atrial sodium channel function and reduces the occurrence of AF.

Rose is excited about the potential of his work.

“We are looking at novel potential treatments and may be able to take advantage of existing compounds to prevent the remodelling that takes place in the heart leading to AF in people with diabetes,” says Rose. “This is an exciting area of study.”

Other grants

Dr. David Campbell, MD, PhD, and colleagues Drs. Dana Olstad, PhD, Eldon Spackman, PhD, Reed Beall, PhD, and Lorraine Lipscombe, MD, received $430,908 for their project, “Healthy Food Rx - Empowering Food Insecure Albertans to Manage Their Diabetes Through a Subsidized Healthy Food Prescription Program: A Randomized Controlled Trial and Economic Evaluation.”

This funding will allow the researchers to expand on their FoodRx trial, which has been funded through Partnership for Research and Innovation in the Health System (PRIHS-5) and a CIHR priority announcement funding. The randomized controlled trial assesses whether individuals with diabetes and food insecurity achieve lower blood sugars when they receive a financial incentive for purchasing healthy foods, compared to those who do not receive the incentive.

The goal of the project is to help inform policy-makers on whether healthy food prescription programs should be offered more broadly to help patients with diabetes and food insecurity to achieve better blood sugars and lower their risk of complications.

Drs. Dana Olstad, PhD, and Joon Lee, PhD, received $100,000 for their study, “Using Artificial Intelligence to Study the Impact of Digital Marketing of Unhealthy Food on What Children Eat.”

The project involves expanding an artificial intelligence (AI) system that identifies digital marketing of unhealthy food to capture all the ways marketers try to appeal to children from a child’s perspective. The project will also help researchers understand how policies can better protect children from the negative impacts of unhealthy food marketing.