Sub-Project 2

Validation of Co-Processor Technologies in the Non-Human Primate

OVERVIEW

Before deploying novel recording and stimulation technologies in humans, we must rigorously characterize their performance and effects on the brain. Non-human primates (NHPs) offer the closest biological analog to human brain circuitry, enabling us to systematically compare invasive and non-invasive approaches and optimize stimulation parameters for maximum therapeutic effect.

WHY EPILEPSY PATIENTS?

Epilepsy patients undergoing pre-surgical monitoring provide a rare and invaluable window into the live human brain at millisecond and millimeter resolution. These individuals have electrodes implanted in cortical and subcortical areas for clinical purposes; with their consent, we conduct cognitive experiments during their monitoring period. Their brain activity during tasks provides direct insights into the neural computations that co-processors need to replicate.

RESEARCH GOALS

Compare the performance of invasive recordings (spikes, local field potentials) versus non-invasive EEG for decoding visual and attentional signals

Test and optimize stimulation paradigms — varying frequency, amplitude, and waveform — to identify which approaches best enhance functional connectivity

Build a harmonized dataset of matched invasive and non-invasive recordings from both primates and humans

WHY VISUAL AND ATTENTIONAL DECODING?

The visual domain offers a uniquely controlled experimental setting: inputs can be precisely defined, neural responses are well-characterized, and canonical computations like normalization and gain control are well understood. This makes it ideal for rigorous benchmarking of recording and stimulation technologies before application in the more complex motor domain.

INFRASTRUCTURE

  • Invasive recording arrays (Utah arrays) in visual areas V1, V4, IT, and prefrontal cortex
  • Simultaneous EEG recording during identical task paradigms
  • Transcranial and intracortical stimulation systems for direct comparison
  • Advanced machine learning pipeline for decoding and comparison

CLINICAL PARTNERS

  • Amrita Advanced Centre for Epilepsy (Kochi)
  • MS Ramaiah Memorial Hospitals (Bengaluru)
  • Sree Chitra Tirunal Institute (Thiruvananthapuram)
  • Deenanath Mangeshkar Hospital (Pune)

EXPECTED OUTPUTS

  • Determination of how well visual stimuli and attentional states can be decoded from invasive vs. non-invasive signals
  • Identification of optimal stimulation paradigms for enhancing inter-regional functional connectivity
  • Open data repository and algorithm suite for harmonized primate-human neural decoding