Three Likely Applications of AI in Integrative Healthcare

Over the past five years, the capabilities of artificial intelligence (AI) have rapidly advanced, accomplishing tasks never thought possible by a machine. And while scientists cannot fully replicate the human brain, in many ways, AI has the potential to surpass human abilities, especially in healthcare, according to Greg Elliot, MS, CEP, DO(MP).

“It’s one of those things that’s inevitably going to change the way healthcare is being delivered,” said Elliot, an osteopathic practitioner in Vancouver, Canada and Co-Founder of HealthQb Technologies Inc. “We don't know how long it will take, but we're in the mix of that transition now.”

According to Michael Snyder, PhD, Chair of the Department of Genetics and Director for the Center for Genomics and Personalized Medicine at Stanford University, AI is a way of probing data with machine learning to garner insights and associations not seen through standard measures. In turn, Snyder explained, AI can identify patterns in a patient’s health metrics that the human eye cannot, which could be helpful in many healthcare domains, particularly diagnostics and disease prevention.

Uses for AI in Integrative Medicine

To Snyder, integrative medicine and AI technologies go hand in hand. Unlike conventional medicine, integrative care focuses on the individual with the acknowledgment that each body is unique and has different reactions to the same diseases and treatments. Snyder said AI quantifies those differences, so doctors can better understand the problem and tailor their treatment plans accordingly.

“Healthcare is all about data; you need data to assess what your health system looks like, and you need it to make recommendations on how you should act,” said Snyder. “The reason this is such a big deal is that we're all different, and we have individual baselines. You can't just lump people into one box and treat them one size fits all.”

Elliot agreed, explaining that AI provides a more holistic and objective view of a patient’s health, making way for a truly personalized treatment plan. For integrative medicine, Elliot said there are three main applications of AI, including:

1. Behavioral Change

One of AI's most likely clinical uses will be to support patients in changing their lifestyle behavior, explained Elliot, who has already incorporated conversational avatar chatbots in his practice as an educational resource for patients.

AI chatbots can answer basic questions about symptoms and medications. And when the chatbots are asked to help with a situation outside of their capabilities, they can instruct patients to contact their provider, Elliot said. This technology increases efficiency, he explained, allowing patients to feel more support outside of an integrative practice while minimizing their correspondence with providers over basic questions.

Paired with wearables data, Elliot said AI technology can also provide patients with insights into their health and make recommendations based on them.

“We integrate with a bunch of wearable data information targeted to understand intervention efficacy,” explained Elliot. “Based on the information that comes through, AI insights can show if things are going well or not so well.”

Overall, Elliot said integrative practitioners should view AI technology as a supportive tool to help reinforce the interventions they recommend to patients.

“It’s important to accept this type of thing not as a replacement but more as a support tool in many different ways,” said Elliot. “The main goal is to improve patient outcomes, and I think the use of AI is absolutely going to be able to help with that.”

2. Predictive Technology

To Snyder, among the most revolutionary aspects of AI in healthcare is the ability to track data longitudinally. This means that scientists will be able to establish a person’s baseline health metrics and track how they change over time to detect any abnormalities.

“In the future, AI will be tracking your health kind of like a rocket ship, looking for things that shift off course," said Snyder.

In addition to predicting health outcomes based on patients’ health metrics, AI can also identify previously unknown markers of disease through genome analysis. For example, Snyder and his colleagues at Stanford identified 683 new genes associated with amyotrophic lateral sclerosis (ALS) with AI technology. Before Snyder's study, scientists only knew of seven ALS-related genes.

According to Elliot, AI will be extremely valuable in spotting and monitoring disease markers like the genes associated with ALS identified by Snyder, which will allow for earlier interventions to prevent disease progression. 

“By having more immediate feedback from individuals and getting data from various technologies, we can start to look at warning signs and detect certain things that need to be addressed,” said Elliot. “Then we can flag at-risk individuals to seek medical care.”

3. Diagnostics

According to Elliot, another future application of AI will be in diagnostics to make the process more automated. However, Elliot said it will be a long time before the technology can diagnose certain conditions without a doctor's input.

“Diagnostics is definitely one of those things that could be in the realm of possibilities as we go forward and learn what these machines can do,” he said.

Understandably, incorporating AI into the healthcare model has been met with resistance from many in the medical field. AI systems are still being developed, and the information they take in is not always scientifically accurate, explained Elliot. Without human involvement, there’s a significant risk of AI machines making jump decisions based on a non-holistic picture of a patient, causing considerable damage.

“Making those concrete decisions based on clinical judgments, it’s not there,” said Elliot. “But as things continue to grow and move, I have an indication that it will get there, and there will be less oversight from human beings and more oversight from artificial intelligence.”

Incorporating AI into Your Practice

Elliot said that the best way to learn about AI is to start playing around with large language models. “Ask the AI about various questions and symptoms and just see what it can possibly do and the limitations of it all,” he explained.

Snyder also believes it’s important for practitioners to become comfortable with AI. Despite the concerns associated with AI in healthcare, Snyder said the technology is not going away and that in the end, it will improve medical technology and health management, not set it back. “Think of cars,” he said. “The worst new car today is one hundred times more manageable than one from 30 years ago.”

Elliot said AI is the future of healthcare, and providers should get ahead of the curve and embrace the technology. “We can’t ignore this transition from the way things were done to the way things will be done,” he said.