Researchers have created a machine-learning algorithm called SELSER that analyzes EEG data to determine whether a patient is likely to respond well to the popular antidepressant called sertraline, according to the National Institute of Health. The algorithm works by looking for a particular neural signature involving complex brain activity patterns linked to positive outcomes from taking this medication.
Clinical depression is a common mental health condition that is difficult to treat. Though there are many different types of antidepressants on the market, the most commonly prescribed variety is SSRIs, of which sertraline is one of the most popular options. While some patients respond well to this medication, others do not experience an improvement in their depression symptoms and may, in fact, feel worse on the medication.
Scientists have identified a neural signature that is linked to positive outcomes from sertraline and trained a machine to identify this signature in patients. The technology may one day help doctors determine whether prescribing this SSRI is likely to help the patient based on their brain activity patterns.
This is a welcome alternative to the trial-and-error approach currently used by doctors to determine the best option for treating a particular patient’s depression. According to the study, the SELSER algorithm ‘reliably’ predicted how participants’ would respond to sertraline based on their EEG info.
As well, the algorithm could potentially also predict ‘broader clinical outcomes’ beyond just how well the patient would respond to this SSRI. For example, the algorithm predicted that patients who didn’t respond well to sertraline were more likely to respond to transcranial magnetic stimulation and psychotherapy.