The computer learned to predict the effectiveness of antidepressants on EEG

American scientists have developed an algorithm that can predict the effectiveness of the course of antidepressant of sertraline on brain activity at rest. To do this, they trained a regression model on EEG 309 patients, half of whom were prescribed a course of the drug, and half placebo. Was that the best way to predict the effectiveness of treatment allows activity at rest with eyes open in the alpha range and can be used to test other therapies, for example, magnetic stimulation. Article published in Nature Biotechnology.

At the molecular level depression is the lack of neurotransmitters from the group of monoamines, which are primarily responsible for the transmission of an electrical signal in the reward system of the brain: serotonin, dopamine and norepinephrine. In most cases, for the treatment of depression using drugs from the group of selective inhibitors of serotonin reuptake: their efficiency they showed in the middle of last century.

However, drugs in this group very much, and they differ in various additional effects, such as antianxiety. Due to the fact that estimate all additional symptoms and choose the right medicine is difficult, the effectiveness of antidepressants often, even in clinical studies is not very high (namely, not higher than placebo): just because the drug is not suitable.

Therefore, in order for the treatment had at least some effect, whether the drug to act, it is better to know in advance. One way is to study the effect of the drug on the brain: for example, at the end of last year, scientists managed to predict the effectiveness of several different antidepressants for relations of the brain.

To assess the impact of treatment on brain activity seems more meaningful, at least in the short term: the patterns of brain activity should give in to treatment faster than its connection. Scientists under the leadership of Atkin, Amit (Amit Etkin) from Stanford University decided to evaluate the effectiveness of antidepressants on EEG at rest.

Their study involved 309 people diagnosed with depression. Participants were divided into two groups: the first within eight weeks of taking a course of antidepressant of the sertraline (it is also known under the brand name “zoloft”), and the second — a placebo. Before the course, scientists have removed the indicators of brain activity at rest with closed and open eyes. Upon completion of the course the effectiveness of the treatment was evaluated by Hamilton scale, compared with the figures before treatment.

The scientists focused on the activity of prefrontal and parietal areas of the brain and trained on the data obtained, the algorithm based on linear regression it is possible to highlight the relationship between treatment efficacy and brain activity at rest. For the predictions used the activity in four main bands: theta (4 to 7 Hz), alpha (8 to 12 Hz), beta (13 to 30 Hz) and gamma (30 to 51 Hertz).

For the group that received sertraline, the effectiveness of treatment is best predicted by activity in the alpha band at rest with eyes open (p < 0.0001), but the effectiveness of placebo, it could not be determined. In this case the algorithm found no relationship with the activity and severity of depression symptoms at the initial diagnosis.

The algorithm is then tested on two additional independent samples of patients (24 and 72), and the effectiveness of treatment with sertraline was also able to predict prior to treatment according to EEG. After this, the researchers used data obtained from 152 patients depression who were treated with transcranial magnetic stimulation: it was found that the algorithm can also predict the efficiency rate of stimulation of the dorsolateral part of the prefrontal cortex.

Scientists, therefore, not only managed to develop a rather simple method of analysis of the effectiveness of antidepressants, but to point to a separate neural mechanism of activity in the alpha band at rest — by which to evaluate other methods of treating depression.

Efficiency is another common method of treating depression — psychotherapy — may also be measured via the EEG, but by how the brain responds to reward, you can even predict what is psychotherapy or antidepressants would work better. Two years ago learned to do American psychologists.

Leave a Reply

Your email address will not be published.