NEW BRUNSWICK, N.J. — Precision medicine is a promising new technology-based approach to health care, but researchers from Rutgers University say there’s no one-size-fits-all AI program currently capable of covering the full spectrum of human health.
Study authors analyzed dozens of artificial intelligence programs used in precision (personalized) medicine to prevent, diagnose, and treat various diseases. They ultimately concluded that no single program exists which is capable of handling all medical situations.
What exactly is precision medicine?
Precision medicine is an AI-based approach that’s all about focusing on the individual patient’s medical history and genetic profile, relating that information with the medical experiences of others, and identifying patterns that will help prevent, diagnose, and treat disease. Due to the sheer volume of medical and genetic information that these programs need to analyze to find patterns, this type of strategy requires an incredible amount of both computing power and machine-learning intelligence.
“Precision medicine is one of the most trending subjects in basic and medical science today,” says study leader Zeeshan Ahmed, an assistant professor of medicine at Rutgers Robert Wood Johnson Medical School, in a university release. “Major reasons include its potential to provide predictive diagnostics and personalized treatment to variable known and rare disorders. However, until now, there has been very little effort exerted in organizing and understanding the many computing approaches to this field. We want to pave the way for a new data-centric era of discovery in health care.”
Current AI systems suffer from too much disorganization
This study zeroed in on 32 of the most prevalent programs currently in use to study preventive treatments across a broad array of diseases — including obesity, dementia, breast cancer, major depressive disorder, and inflammatory bowel disease. The research team combed through five years’ worth of high-quality medical literature to identify these programs. They say that while it’s clear the greater precision medicine field is advancing at a rapid pace, it’s also quite disorganized.
AI programs simulate human intelligence processes. Machine learning, meanwhile, is a subcategory of artificial intelligence; these programs “learn” as they continue to process new data, gradually becoming more and more accurate at predicting outcomes. The actual work falls on the algorithms, or step-by-step procedures for solving problems or performing computations.
According to Prof. Ahmed, the use of genetics is “arguably the most data-rich and complex component of precision medicine.” With this in mind, his team placed particular importance on reviewing and comparing scientific objectives, methodologies, data sources, ethics, and gaps used by the programs.
In conclusion, study authors say the scientific community needs to embrace several “grand challenges” to help facilitate the success and adoption of precision medicine. These challenges range from fixing general issues such as improving data standardization or enhancing the protection of personal identifying information to more technical problems including the correction of either genomic or clinical data.
“AI has the potential to play a vital role to achieve significant improvements in providing better individualized and population healthcare at lower costs,” Prof. Ahmed adds. “We need to strive to address possible challenges that continue to slow the advancements of this breakthrough treatment approach.”
The findings appear in the journal Briefings in Bioinformatics.