How is AI making a difference to precision medicine?
Precision medicine is undoubtedly the phrase on everyone’s lips, as we eagerly anticipate the next breakthroughs in medical technology, disease prevention, diagnosis and treatment. Researchers at Weill Cornell Medicine in Qatar have outlined where the greatest opportunities and risks lie, against the backdrop of artificial intelligence (AI).
The first point to note is that precision medicine still has a long way to go before it begins to realise its full potential. A recent Economist report, for example, highlights the challenges of data management as key (how to manage, harness, understand and apply enormous volumes of data). AI (including machine learning and deep learning) come into their own when analysis of ‘big data’ is needed and offers unparalleled support to the development of precision medicine. The Will Cornell authors also cite communication tools and wearable technology as another opportunity to support precision medicine.
Faster collection of genomic data is another development that has given a boost to precision medicine. With faster and cheaper genomic analysis technologies increasingly available. With the growth of databases of genetic data, and the identification of markers of specific conditions within them, and by linking that data to specific populations and cultural groups, the path towards true precision medicine is becoming clearer. For example, the Qatar Biobank is seeking “to reduce the number of chronic illnesses … through medical research on the causes of prevalent illnesses.” Looking to the future, Qatar Biobank hopes its expanding data collection and analysis will “reveal the effect of lifestyle, environment and genetics on Qatari population.”
The COVID-19 pandemic has given some insight into how these tools might develop precision medicine approaches further. An AI-driven approach to medicine development can also shed light on how to repurpose existing treatments, guiding decision-making. AI guided doctors developing treatment plans for COVID-19 allowing them to classify infection severity and point towards most effective treatments.
The future will likely see better and more accurate outputs from AI – as the volume and quality of data that we feed into such systems develops. Some recent AI cases have illustrated how it’s still very necessary to have expert humans overseeing the outputs, though the successes AI has delivered already can be built upon, particularly in the areas of rare and genetic diseases.