Big data: An algorithm for profound change in heart health

May 14, 2019

Canadians are living amid an explosion of information and innovation thanks to significant advances in data science and technology. Capable of seeing patterns in big data that would otherwise remain undetectable by humans, finding the proverbial needle in the haystack is now not only possible but practical. With this new computational capacity comes the potential to revolutionize health care and a promising pathway forward for the future of cardiovascular medicine.

As human beings, we’ve been gathering and analyzing data to improve our lives and societies since we first started taking stock of domesticated animals and crop yields. But, in the modern digital age, massive amounts of complex data are so deeply embedded into the very fabric of our lives that information itself has become the dominant currency of our societies. We endeavour to employ rich evidence-based data in almost every sphere of life to enhance our well-being, address pressing societal issues and drive economic prosperity.

Once the sole domain of governments and tech titans like Google to foster insights into their constituents and customers, big data analytics – artificial intelligence (AI) and machine learning – are now increasingly being used by heart researchers, clinicians and other health innovators to improve patient healthcare and outcomes.  

“There is no doubt that machine learning, and the big data and the AI that empower it, is promising to revolutionize the health care sector,” says Marc Lamoureux, Manager of the Digital Health Division in Health Canada’s Medical Devices Bureau. 

Lamoureux was one of the many subject matter experts highlighting leading data science technologies in cardiovascular disease at the 7th International Ottawa Heart Conference: Big Data in Cardiovascular Disease, jointly held with the Toronto-Ottawa Heart Summit recently in Ottawa. Through his work at Health Canada, he is seeing the emergence of machine learning in image-based healthcare applications such as diagnostic imaging.

“Anomaly detection and image classification are just two of the applications of AI in diagnostic imaging,” he says. “We are just at the doorstep, but the potential of big data and AI in advancing cardiac diagnosis, disease onset prediction, prognosis, and more, is quite staggering.”

What is new about big data?

In simple terms, big data is essentially data that exists on a massive, and massively complex scale. It is chaotic and full of noise, and its trajectory seems unbounded since the advent of the Internet, social media networks and sensors embedded in devices like wearable health technology and smartphones.

This information has existed all around us for some time now. What is new about big data are the technological advancements in computing power and storage capacity that have allowed us to capture and process the ever-growing volume, value, veracity, velocity, and variety – the so-called “5 V’s” – of data generated today.

What is new and exciting about big data in cardiovascular medicine is our ability to use AI to unpack the richness of medical data.

- Dr. Peter Liu, Chief Scientific Officer and Vice President of Research at the Ottawa Heart Institute

Dr. Liu joined the Heart Institute as the scientific director in 2012. He received his M.D. and completed his postgraduate training in internal medicine and cardiology at University of Toronto and Harvard Medical School.

"What is new and exciting about big data in cardiovascular medicine is our ability to use AI to unpack the richness of medical data,” says Dr. Peter Liu, Chief Scientific Officer and Vice President of Research at the Ottawa Heart Institute.

“Now we can integrate this information with new data from a patient’s smartphone sensors or wearable health device to help us develop important insights into how to provide optimized healthcare,” he adds. “The living patient and the environment in which that patient lives, with all their different medical parameters, can become part of a patient’s repertoire.”

The emerging role of big data and AI in cardiovascular medicine

As the leading cause of death and disability globally, cardiovascular disease (CVD) remains a heavy burden. Complicating matters is the fact that CVD doesn’t affect everyone in the same way. A person’s vulnerability to the disease and responses to treatment stem from a highly complex interaction between genetic predispositions and modifiable environmental factors such as exercise behaviour, diet, and smoking history, among others.

Until very recently, researchers and clinicians in cardiovascular medicine couldn’t hope to integrate all these different factors when diagnosing a patient or devising an appropriate treatment plan. In the era of digital health and big data, the generation of data is so large and complex that it has become impossible for human beings to meaningfully manage all of it using traditional databases and analytic tools.

Enter artificial intelligence.

“If the industrial revolution was about machines extending humans’ mechanical power, the AI revolution is about machines extending humans’ cognitive power,” explains Dr. Doina Precup, Canada-CIFAR AI chair at McGill University/Mila national AI Institute and Senior Member with the American Association for Artificial Intelligence.

Although AI is still in its relative infancy in cardiovascular medicine, Dr. Precup believes that it can improve lives around the world by facilitating early disease detection and prognosis, better access to top-notch health care, and by enhancing our own cognitive abilities.

Already, deep learning AI algorithms have been shown to make diagnoses at least as well as physicians in cardiology and other fields – so does this mean that AI will eventually replace flesh and blood doctors?

“Absolutely not,” says Henry Lieberman of the MIT Computer Science and AI Lab (CSAIL) who says that while AI holds great potential for helping cardiologists with diagnoses and workflow and patients with wearable tech to monitor their hearts, it will never replace the human touch.

“AI is great at finding patterns in data, but it’s no substitute for human common sense,” he says. “We will need a tight integration of physicians and AI practitioners to continuously interpret and contextualize its applications in clinical cardiovascular medicine.”

Clearly, no one is going to be able to realize the power and promise of big data and AI applications in cardiovascular medicine – or any field for that matter – on their own. The challenge of translating the inordinately complex and exponentially growing volumes of big data for clinical practice will take a shared and immense effort.

“The new tools we are using, and the new data that we are generating, can lead us to achieve profoundly better patient-centric care and health,” says Dr. Liu. “But it will only be through communication, through sharing, and through mutual support locally, across the country and internationally, that we realize this goal.”

Under the pressure of explosive growth of both data and technology, healthcare will change. For researchers and clinicians in cardiovascular medicine, big data and AI are unquestionably helping to write a new chapter. Have your reading glasses ready, it’s going to be an exciting read.