Machine Learning and Healthcare: The Evolution of Care

By Lester Mcpherson, National Director – Customer Acquisition, Success and Expansion at Strata Health Canada 

We’re at the beginning of 2024, and one thing’s for sure: Healthcare has been forever changed by the possibilities brought about by a mostly mysterious term – “Machine Learning” or what marketing promotes as “AI.” 

Much like ice cream, there’s a variety of flavours to tickle one’s fancy: For those less adventurous there’s Reactive AI that has a pre-determined set of responses based on scenarios that are encountered. The scale travels all the way up to Deep Learning where there’s more of a “set it and forget it” ability as the system will actually learn, and associate behaviors based on classifications of stimuli within scenarios. It’s a bit difficult to imagine how AI can fit into our world, and in particular healthcare so let’s look at 3 trends currently active in healthcare.  

Number 3: Clinical Analysis & Staff Resourcing 

This one is particularly interesting in Canada.  It’s well known that the health care system in Canada is besieged with long wait times. It’s not uncommon for folks to avoid ERs due to long waits or accept that a specialist appointment needed urgently can take months. The Financial Accountability Office (FAO), which is a body that provides independent analysis on the state of Ontario’s finances, estimated that there will be a shortage of 33,000 nurses and personal support workers by 2027. 

AI especially in rural areas will be having a huge impact as on average there is less staffing in general within health care organizations. The CBC recently highlighted this with a quote from Dr. Alex Wong, the Canadian Research Chair for Artificial Intelligence. Dr Wong said, “When you see a doctor on a computer, they’re looking at images, records and data. Now you have this additional AI that provides additional insights and information.” Clinical analysis assisted by AI effectively speeds up all operational aspects of clinical operations, including predicting trends in-patient admissions, identifying potential outbreaks of diseases, and reducing errors. 

This expedited form of analysis allows for more patients to receive higher quality care. Many people have experienced “missed diagnoses” and AI is the first step in reducing this.  

Number 2: Diagnostic Imaging 

Nothing is worse than holding your breath while waiting for results from a scan – most of us have been there unfortunately. Sometimes we feel like we are at the mercy and capability of the specialist to identify something that could be very rare, or something that is very time sensitive. 

Many organizations are proactive in evolving their practice to include AI. For example, the Canadian Association of Radiologists, while recognizing the need for oversight, is embracing AI. The Radiology AI Validation Network is an assembly of AI specialists in the field of radiology tasked with assessing the viability and performance of AI applications.  

Now here’s the exciting part. Looking to our friends in the U.S, the U.S. Food and Drug Administration (FDA) can only approve algorithms involving imaging that meets the analytical threshold of being accurate 80 to 90 percent of the time. Last year, the FDA approved 420 of these, which is a great start all things considered. Taking into account that AI’s use and the understanding of how it can be applied grows exponentially, we can reasonably expect that there will be 1000’s of algorithms at work by 2030. Why is this exciting? It’s because diagnosis, such as Cancer, requires the monitoring of patients in tracking even the smallest changes. This is now not only more efficiently tracked, but also predictive trends allow for health professionals to significantly increase rates of survival through early detection and prevention, as well as newly designed proactive procedures, which leads us into Number 1: Personalized Healthcare. 

Number 1: Personalized Healthcare 

This, in my opinion, is the finishing piece in the healthcare revolution that AI is bringing about. Every one of us has a hereditary disposition toward various ailments. Additionally, one size has never fit all in health care; many scenarios are “best efforts” by health organizations based on information from previous medical science. 

Imagine an AI system that analyzes your bloodwork, diagnostic imaging, and previous medical history. Now imagine a system that continues to monitor your “health levels” through your smart device, such as a watch or phone, etc.  

When we talk about the previous points 2 and 3 earlier in this article, their outputs enable AI to deliver a level of personalization in medicine that was previously unattainable. Treatments and recommendations can be tailored to the individual’s genetic makeup, lifestyle, and environmental factors. AI in the future will connect you to telemedicine services based on your current diagnosis – much of the guess work will be significantly reduced.  

No Sunshine without Rain…. 

It would be irresponsible to celebrate AI’s potential without touching upon important challenges. Bias, for example, could have AI interpreting a patient less impartially due to gender or other genetic aspects, for example. Ethics debates are heavy right now on various topics around consent and data privacy to name just two; since AI is based on data models need a voracious amount of data to continue to be relevant. The more individuals that give content to share their data makes the system better overall and allows for a greater depth in the AI’s understanding of human physio and mental structure. 

Where do we go from here? 

Healthcare has previously been a sole journey. If AI is to fully deliver a utopian future where disease is significantly reduced if not irradicated in many instances, it will take a shift in society’s idea of the ‘sole’ journey. 

The journey needs to become a collective one. Rather than “How can medicine make me better” it needs to be “How can my DNA / Health Information” make healthcare better for everyone in general? AI is the first real chance in healthcare for humanity to come together and leave none behind. 

---