21:00, 25 November 2025

In Russia, Parkinson’s Disease May Be Defeated by “Smart” Brain Stimulation

Russian researchers have introduced the world’s first universal software environment for training and testing adaptive deep brain stimulation algorithms, paving the way for safer and more personalized therapies for Parkinson’s disease and other neurological disorders

Russian scientists have unveiled the first global software framework designed to train and evaluate algorithms for adaptive deep brain stimulation (DBS). The platform is expected to accelerate the development of safer and more effective treatment strategies for Parkinson’s disease by fine‑tuning stimulation in real time to reduce side effects while maximizing symptom suppression.

Deep brain stimulation is often prescribed when medication no longer provides relief. Electrodes are implanted into the basal ganglia, a region where Parkinson’s patients exhibit abnormal neural activity that leads to motor impairments. Electrical stimulation suppresses this activity, easing rigidity and hand tremors. However, traditional DBS operates continuously, and even low‑intensity, nonstop stimulation can produce side effects, including speech disturbances, and may gradually affect neural tissue.

The new approach changes the paradigm. Modern electrodes are capable not only of stimulating neural tissue but also of recording brain activity. This enables adaptive DBS — a closed‑loop system that adjusts stimulation based on real‑time neural signals. Until now, researchers lacked a unified environment for developing and comparing such control algorithms. The new Russian platform fills that gap.

“We created a software environment where adaptive control algorithms — whether AI‑driven or more traditional — can be tested,” said project lead Dmitry Dylov of Skoltech and the AIRI Institute. “The system incorporates a model that simulates how neurons respond to stimulation, how brain activity evolves, and even how neural behavior changes over time.”

The platform can model various patient states, including sleep and walking, enabling artificial intelligence algorithms to be trained on highly realistic synthetic data.

“Bidirectional neural interfaces — those that read neural signals and simultaneously control neurostimulation — represent the cutting edge of neurotechnology,” noted Mikhail Lebedev, a professor at Moscow State University. “Practical systems already exist, but we still lack a full understanding of how they should work. Our research brings us closer to that understanding. While the current focus is Parkinson’s disease, the implications extend to other neurological disorders as well.”

The introduction of this platform opens new possibilities for the global neuroscience community, giving researchers a powerful and flexible tool for testing next‑generation therapies.

Medicine and healthcare