This SciShow video is supported by Ground News, a website and app that lets you compare how major events are being covered so you can see more sides of more stories.

You can go to ground.news/scishow or click the link in the description to get 30% off the Vantage level subscription. Show of hands, who here has sought treatment for depression?

Ok, now how many of you got your meds right first try? Yeah, I didn’t think so… Many people know the painful, back and forth, months long process of cycling through meds, trying to find the right one. This process might be soon get a lot simpler by using brain scans to predict which meds are just right for each person. [♪ INTRO] Because everyone experiences depression differently, it can be very difficult to treat consistently.

Right now, doctors use a pretty blunt trial and error approach. A kind of see what sticks deal. You might go through six rounds of meds, each 3 months per round, before finding something that works.

So a tool to effectively predict how a patient will respond to an antidepressant before they even start it is something doctors and patients alike would really like to see. And that tool might finally be right around the corner, in the form of an electroencephalogram or EEG. An EEG is a test that uses electrodes placed on the scalp to record the electrical activity of neurons in the brain.

Now, it’s not like people have never thought to use EEGs to predict responses to antidepressants before. But these studies often produce conflicting results due to small numbers of participants, and just how different depression is between people. A 2023 study might have found a way to break through all the noise by using machine learning to interpret the EEG results of a much larger group of people.

The study pulled data from two large EEG datasets collected by CAN-BIND and EMBARC, large research efforts trying to find biological markers of depression that might help match a patient to the right meds. Markers like brain activity. Participants in both groups received meds from a class of commonly used antidepressants called SSRIs.

The CAN-BIND group received escitalopram while the EMBARC group received sertraline or a placebo. First, the researchers fed the EEG data from participants in the CAN-BIND study into a computer model so it could learn to predict how people would respond to escitalopram based on their distinct electrical brain activity. With the model ready to go, they used it to analyze the EEG data of participants in the EMBARC study to see if they could predict how they would respond to sertraline.

In the end, the model was able to predict the response to both drugs with over 60% accuracy. I know what you’re thinking and, yeah, 60% may not sound like much – but it creates a starting point that isn’t a complete shot in the dark! Now, this still isn’t perfect – and the authors of the study acknowledge that.

Analysis of more datasets on larger groups of participants will likely be needed to refine the model and make it even more accurate. And the types of meds included in these future studies will need to be carefully considered. For example, it would probably be more helpful to create models that let us compare different classes of antidepressants, rather than two flavors of the same class.

Despite that, this is a really promising start. The researchers intentionally set up their model so that it could be used with cheapo EEGs. A lot of clinical research uses swanky EEGs that have more electrodes and require more specialized training.

But this model is meant to emulate what your physician might have access to – making it that much more feasible in the real world. I can’t make any promises, but this could be coming to a doctor’s office near you. Thanks to Ground News for supporting this SciShow video!

When it comes to topics like brain scans, medication, or science, news outlets can ignore certain stories based on their narrative. But it’s important to get different viewpoints and stay well-informed of our blindspots to make sure we’re consuming and sharing the most accurate information. That’s where Ground News comes in.

Ground News is a startup founded in 2018 by former NASA engineer Harleen Kaur. Their app and website compile news from around the world into one place, offering curious people like you and me a comprehensive view of current events. For example, they compiled over 35 sources reporting on the ‘’First Postpartum Depression Pill Now Made Available in the US’’.

The majority are from politically centered publications, with none coming from any right-leaning sources. So this big development might be missed by anyone relying only on right leaning sources. And with Ground News, you can even identify which states are and aren’t reporting on this topic.

You can look into it yourself by visiting ground.news/scishow or clicking the link in the description below to get 30% off unlimited access. Thanks for watching SciShow! [♪ OUTRO]