He had shown in 2006 that some patients with different diagnoses — PTSD vs. depression, for example — looked remarkably similar under brain imaging, suggesting clinicians drew distinctions in the wrong places. And in 2014, he showed that one could define patients by looking at individual discrete behaviors, such as attention or sleep.
But the big advance came in 2017, when his team demonstrated it could predict PTSD patients’ response to therapy by just looking at their brainwaves. Similar predictions, aided with a bit of AI, followed for other disorders.
“So those are the tools that really got us excited,” Etkin told Endpoints News. “And then of course, the main question that you can’t answer in the lab is, how do you develop new therapeutics around this?”