Deon Ramgoolam: The algorithm will see you now—and it might save your life

For decades, diagnosing heart disease has relied on a mix of human skill, sophisticated imaging, and a bit of detective work. But now, Canadian clinics are adding something new to the equation: algorithms. The latest wave of AI diagnostic tools, such as Us2.ai and DeepECG.ai, can parse heart data in minutes, spot subtle abnormalities that humans might miss, and deliver consistent results anywhere in the country. While this technology is not a replacement for medical judgment, it can act as a force multiplier. This is especially important given how critical it is to find heart disease before it turns into a crisis. AI in the echo lab: Us2.ai’s instant insight When a cardiologist examines the heart using an echocardiogram, they interpret waveforms, measure pumping power, and calculate a measurement called global longitudinal strain (GLS). These steps often take longer than 45 minutes to complete. Yet Us2.ai has reduced that to under two minutes. It’s extremely fast, guideline-driven, and reproducible, with no scribbled measurements required. Even more strikingly, it can detect stealthy threats like cardiac amyloidosis from a single echo view, now cleared by both the FDA and CE regulators. In essence, Us2.ai brings robot-grade precision to everyday echoes. From lines to life: DeepECG.ai’s Canadian connection Twelve ECG leads, once the domain of crumpled paper and hurried scribbles, have become a playground for DeepECG.ai. Feed it an ECG trace and, within seconds, it can flag atrial fibrillation, left ventricular hypertrophy, or even signs of an impending heart attack with near-perfect accuracy. The true charm of it? This isn’t happening somewhere else, but in Canada, at the Montréal Heart Institute, under clinical trials such as DAISEA-ECG and HEART-AI. These studies are exploring how AI-assisted ECGs can help doctors detect heart disease earlier and reduce diagnosis times. The result could be fewer misdiagnoses and fewer delays. The cost of what’s missed: Why early detection matters Heart disease also drives hidden costs that never appear on a balance sheet. These include informal caregiving, where family members reduce work hours or leave jobs entirely to care for loved ones, and the cascading effect of disability on household income. Studies from other G7 nations suggest that each case of advanced, undiagnosed cardiovascular disease can drain an additional $20,000 to $40,000 per year in indirect costs, a figure that could translate to billions annually in Canada. When multiplied across the estimated 750,000 Canadians living with undiagnosed heart conditions, the economic toll becomes staggering. Beyond being a personal health problem, undiagnosed or delayed heart disease hits our wallets, too. Heart disease already costs Canada an estimated $21.2 billion annually in direct medical care and lost earnings. That’s more than any other disease and includes the effects of strokes and other vascular conditions. For heart failure alone, hospitalizations are projected to reach a staggering $19.5 billion between 2019 and 2040—just for inpatient care. Those figures multiply when you factor in the cost of undiagnosed or poorly managed conditions: extra visits, emergency admissions, lost workdays, and the human cost of families in crisis. AI tools like Us2.ai and DeepECG.ai can mitigate those costs by catching warning signs early, even before symptoms appear. Why these tools matter in Canada In provinces like Newfoundland and Labrador, or in northern regions of British Columbia, travel to see a specialist can involve flights, ferries, and days off work. AI tools deployed in local clinics can close that gap, producing specialist-grade reports on site and allowing remote cardiologists to review them instantly. Communities that have long been at the periphery of specialized care can now be at the centre of timely, accurate diagnosis. Navigating the bumps Integrating AI into Canada’s health system will also mean grappling with interoperability between provincial EMR systems, many of which are decades old. There’s also the matter of clinician trust: doctors must understand how the algorithms reach their conclusions to feel confident acting on them. Pilot programs, such as those at the Montréal Heart Institute, are vital for building trust, allowing clinicians to test AI outputs against their own assessments before widespread adoption. Of course, AI isn’t flawless. Diverse patient populations need validation; clinicians need training; health systems need secure integration. The Canadian projects testing DeepECG and Us2.ai are crucial first steps, but scaling across provinces will require funding, regulation, and buy-in. Consider a 60-year-old man in rural Saskatchewan, experiencing occasional fatigue but no chest pain. In the past, he might have waited months for a cardiology referral and an in-person echocardiogram. Instead, his local clinic runs an ECG through DeepECG.ai, which flags signs of left ventricular hypertrophy. The same day, he gets an AI-assisted echo analyzed by Us2.ai, confirming the abnormality. Within a week, he’s on the right medication and lifestyle plan—avoiding what could have been a costly, life-threatening hospitalization. This is the kind of early intervention AI makes possible, not in theory, but in today’s Canada. A future together Over time, as these systems learn from Canadian patient data, their accuracy will improve even further. The more diverse and representative the dataset, the better these tools will perform for all Canadians—across ages, ethnic backgrounds, and co-existing conditions. If supported by clear policy, robust privacy safeguards, and sustained funding, AI could become as standard in cardiac diagnostics as the stethoscope once was. The promise here isn’t about replacing doctors. But if we can help them make faster, sharper decisions and be better informed, why shouldn’t we? AI tools like Us2.ai and DeepECG.ai can shift diagnosis from reactive to proactive, helping more Canadians avoid the hospital bed altogether. That means healthier lives for patients, and real economic dividends for a health-care system under constant strain. Generative AI assisted in the writing of this article