The Impact of AI in Medicine is Already Noticable

The Impact of AI in Medicine is Already Noticable

The concept that AI may revolutionise medicine may now seem laughable after so many failed attempts. The clinical efficacy of AI, however, is no laughing matter for the family of a six-month-old child who suffers from a seizure problem.


In this instance, researchers and medical professionals were involved in a specific genomics programme at Stanford University that was created to promptly identify rare disorders in newborns. They indicated a baby who was having seizures. They utilised a form of AI known as deep learning, which can be trained to find intricate patterns in vast volumes of data, to compare and analyse the DNA results after quickly sequencing the baby's and parents' genomes.


Using this method, they were able to quickly narrow down the disease's potential genetic causes to a single mutation and establish that it was the cause of the symptoms. Most impressively, it was completed in only eight hours.


According to one of those researchers, cardiologist and genetics expert Euan Ashley, their diagnosis allowed "the bedside team to direct patient care according to the molecular cause of the seizures," according to a recent publication revealing AI's usage in medicine published in The New England Journal of Medicine.


Ashley reports on a number of ways that AI is already making a difference, both for research that will have therapeutic ramifications and, in certain cases, for use directly with patients, together with his colleague Bruna Gomes.


None of it like the AI in medicine that we have grown accustomed to seeing in sci-fi entertainment. No tricorder can diagnose anything, and no artificial intelligence can take the role of doctors in patient care. But over time, researchers and medical professionals are discovering ways to use AI's powerful processing and analytical capabilities to enhance patient diagnosis and care.


The simplification of data interpretation for technological platforms that are on the verge of becoming widely used in medicine but are still not quite there is where most of the near-clinical impact of AI is currently visible. One example is genome sequencing; each of our genomes has a tremendous amount of data, and using it at the bedside will require powerful computing tools that can process data far more quickly and correctly than people.


Deep learning and other AI algorithms are already being used to evaluate this data and indicate DNA variants likely to be associated to particular symptoms, according to Ashley and Gomes. This method can reduce interpretation time and give clinicians actionable information soon enough to benefit patients, as was the case with the six-month-old above.


Beyond genomics, AI is being utilised to enhance data analysis in the areas of proteins, metabolites, and methylation—all of which have the potential to be significant biomarkers that could assist customise healthcare for each patient. Each layer represents a massive array of data that could not realistically be included into standard medical care without the aid of AI.


The real value of molecular medicine today is not found in any one layer of data, but rather in the insights that may be gained from combining them. It will require incredibly advanced AI techniques to gather these large data sets, analyse them in great detail, and filter all of those findings down to the small portion that might be helpful for patient treatment. The study claims that although great progress has already been made in this area, there are still several obstacles to overcome. These include data standardisation, the availability of precise benchmarks, and improving accuracy and speed.


The greatest progress has been made in the identification, and in some cases, treatment, of uncommon genetic illnesses, write Ashley and Gomes. "The outcome is a progressively detailed understanding of the molecular trajectory of disease that is now finding application in clinical medicine."