Researchers have developed a machine learning algorithm that uses electroencephalogram (EEG) data to predict whether a selective serotonin reuptake inhibitor (SSRI) is likely to benefit a patient. They published their findings in Nature Biotechnology.
Researchers from the UT Southwestern Medical Center in Dallas, Texas and Stanford University in California have applied for a patent for the algorithm.
The study included more than 300 patients with depression who were randomized to treatment with an SSRI or placebo as part of the EMBARC (Establishing Moderators and Biosignatures of Antidepressant Response for Clinical Care) trial. Using participants’ EEG data, which measured electrical activity in the brain cortex before treatment, researchers developed a machine-learning algorithm that analyzed the information.
The algorithm, they reported, accurately predicted which patients would benefit from an SSRI within 2 months. In addition, further research suggested patients unlikely to improve with an SSRI were likely to benefit from other interventions, including psychotherapy or brain stimulation.
Researchers validated their findings in 3 additional groups of patients.
“This study takes previous research, showing that we can predict who benefits from an antidepressant, and actually brings it to the point of practical utility,” said Amit Etkin, MD, PhD, a psychiatry professor at Stanford University, who worked with UT Southwestern psychiatrist and EMBARC trial leader Madhukar Trivedi, MD, to develop the algorithm.
Looking to the future, they plan to develop an artificial intelligence interface that can be widely integrated with EEG machines and to gain approval from the US Food and Drug Administration.