Description:
2025-009 MiSO: MicroStimulation Optimization for Brain Modulation with Neural Stimulation Technology
Key Highlights
● MiSO leverages convolutional neural networks (CNNs) to predict neural responses to a wide
range of brain stimulation parameters, allowing for precise and efficient optimization of neural
activity.
● Operates in a closed-loop system that adjusts stimulation parameters in real-time based on
immediate feedback from neural activity, enhancing the adaptiveness and effectiveness of neural
modulation.
● Combines data from multiple experimental sessions to build a robust, cohesive training dataset,
overcoming the limitations of single-session data collection, and enhancing the accuracy of
predictions.
● MiSO's optimization framework is applicable across various neurological and psychiatric
conditions, including epilepsy, Parkinson’s disease, depression, and chronic pain. It provides a
foundation for personalized treatment strategies, which are crucial in modern medicine.
● Adjusts stimulation configurations dynamically to modulate neural activity toward specified states,
ensuring accurate targeting and minimizing adverse effects or ineffective stimulation.
Benefit
● Data Integration Across Sessions: The technology incorporates latent space alignment,
merging data from multiple experimental sessions to create a more robust and comprehensive
training dataset. This overcomes the limitations of single-session data collection, improving the
model's accuracy in predicting neural responses.
● Large parameter space optimization: MiSO enables optimization over a much larger
stimulation parameter space than previous methods, allowing testing of 4,560 double-electrode
stimulation patterns compared to just 96 single-electrode patterns.
● AI-Driven Precision: MiSO utilizes convolutional neural networks (CNNs) to predict neural
responses to a wide range of stimulation parameters. This significantly improves the precision of
neural modulation compared to traditional trial-and-error approaches, which are time-consuming
and less effective in optimizing brain activity. The AI-driven optimization allows for more accurate
targeting and personalization of treatment, leading to better patient outcomes.
● Reduced Trial-and-Error in Therapy: Existing brain stimulation techniques often rely on
repeated trials to find the correct stimulation parameters, which can be inefficient and
uncomfortable for patients. MiSO reduces the need for this trial-and-error process by predicting
effective parameters in advance, thereby speeding up treatment and reducing patient discomfort.
Potential Applications
● Healthcare (Neurology & Psychiatry): Optimizes brain stimulation treatments for neurological
disorders (e.g, Parkinson’s, epilepsy) and psychiatric conditions (e.g, depression, OCD).
● Pain Management: Enhances spinal cord stimulation for chronic pain by dynamically adjusting
parameters for better relief.
● Medical Devices: Integrates with brain-computer interfaces and neuroprosthetics, improving
control and responsiveness.
● Pharmaceutical R&D: Supports combination drug-device therapies for enhanced neurological
and psychiatric treatments.
● Neuroscience Research: Facilitates precise modulation in brain function and plasticity studies.
● Consumer Neurotechnology: Enables cognitive enhancement and personalized
neurostimulation for wearables.
● Rehabilitation Medicine: Assists in post-stroke and brain injury recovery by optimizing neural
activity for improved outcomes.
Publication
IP
PCT/US2025/047445
Graphical Illustration