The powerful technology of Artificial Intelligence, with the ability to not only predict outcomes but also guess future patients' probability of having a disease, is a major benefit for precision medicine, but the tool lacks explanation and regulators fail to verify the results.
But the adoption of the revolutionary tool in health care can take the help of a new type of algorithm called "explainable AI" (XAI) that can be understood by humans and help doctors better diagnose why a disease may occur and in what environments it may recur to evaluate the risk of diseases in patients.
The black box aspect of AI is fine in many fields outside of health care because it allows companies to keep their precious algorithms as trade secrets, but for fields such as medicine this approach of AI makes it difficult for doctors and regulators to trust it because they cannot be sure an AI algorithm will properly diagnose patients.
The algorithm of AI is problematic for the Food and Drug Administration (FDA) because deep learning algorithms find connections and patterns without their operators ever understanding which parts of the data are key to the decision. The FDA in its recently issued draft guidance asked doctors to be able to independently verify the basis for the software's recommendations and avoid triggering higher scrutiny as a medical "device".
However, XAI algorithms, being developed for health care applications, will be able to provide justifications for their results and be relatively simple for humans to understand as "precision medicine" or "personalized medicine", according to the US National Library of Medicine, takes into account individual variability in genes, environment, and lifestyle for each person.
Precision medicine – an emerging approach for disease treatment and prevention -- helps physicians determine more personalized treatments for patients and has an individualized approach instead of a blanket approach for all by looking at a patient's genetic history, location, environmental factors, lifestyle, and habits to determine treatment.
A Chilmark Research said precision medicine in order to achieve its full potential must be accompanied by machine learning and AI due to the deep learning technology and ability to analyze large data sets faster than clinicians and medical researchers.
Developed by pharmaceutical company Amplion, a recent software intelligence platform that uses machine learning to deliver insights into pharmaceutical partnerships known as "Dx:Revenue" uses over 34 million data sources from clinical trials, scientific publications, conference abstracts, FDA approved tests, lab tests, and other information to match a test provider's capabilities to pharma's specific needs.
"This is particularly important in cancer, where we're moving away from the one-size-fits-all approach to care toward a more targeted approach," Forbes quoted Amplion CEO Chris Capdevlia as saying. Precision medicine can make treatments more affordable and accessible, and can better focus on privacy, care quality and patient safety, he said.
A recent e-prescription and medication management company survey said 83 percent of patients "would welcome reminders from their physicians about checking blood pressure, completing rehabilitation exercises, taking medications, scheduling follow-up appointments or other similar activities".