Dealing with the brain is incredibly complicated. While we clearly understand how the lungs fill with air or how the heart pumps blood, the electrochemical impulses that travel through our brains remain largely a mystery. How much of this process depends on the materials between neurons? And how interconnected are the different brain structures? The answer isn’t simple, but that hasn’t stopped us. If you’ve ever tried to fix a computer without any computer knowledge, you’ve probably rebooted it or even crashed it. Something similar happens with the brain.
For a long time, we’ve been doing trepanations, lobotomies, electroshock therapy, and other procedures that, while effective in some cases, can seem brutal, like trying to kill flies in a tube. In the same vein, there are deep brain stimulators, pacemaker-like devices that are inserted deep into the nervous system and deliver tiny shocks with surprising effects, such as reducing the symptoms of Parkinson’s disease in some patients, at least for a while. Now, a new study has used artificial intelligence to take these devices to the next level.
The study involved a small number of patients due to the complexity of the treatment and the high cost of the devices, but it yielded encouraging results. Four people who were already receiving conventional deep brain stimulation were initially asked about their most persistent and bothersome symptoms, despite the treatment. Most of those mentioned were involuntary movements or difficulty initiating movement. The experts then performed deep brain stimulation alongside the conventional treatment the patients were already receiving.
After several months of training with the algorithm, the participants went home with the DBS device. A comparison trial was conducted in which conventional treatments were swapped for adaptive DBS. Surprisingly, the researchers were able to halve the symptoms that patients identified as the most bothersome at the start of the study. This was made possible thanks to machine learning, which made it possible to determine when a patient needed more or less electrical stimulation.
“Infuriatingly humble social media buff. Twitter advocate. Writer. Internet nerd.”
