hh-model-inference-from-mea

name

hh-model-inference-from-mea

description

Framework for rapidly inferring Hodgkin-Huxley (HH) biophysical parameters from extracellular multi-electrode array (MEA) measurements using differentiable biophysical simulation and simulation-based inference. Enables precise neurostimulation prediction without invasive intracellular recordings. Validated on macaque retina with 512-electrode array achieving 90.6% accuracy predicting unseen multi-electrode stimulation responses. Use when: fitting HH models from extracellular data, predicting neurostimulation responses, scaling biophysical inference to large neural populations, designing closed-loop neurostimulation protocols, differentiable neuron simulation, simulation-based inference for biophysical models. Trigger words: HH model fitting, extracellular MEA inference, neurostimulation prediction, differentiable biophysical simulation, simulation-based inference, Hodgkin-Huxley parameter inference, MEA-based modeling, biophysical parameter estimation.

Hodgkin-Huxley Model Inference from Extracellular MEA Data

Paper

Title: Learning Biophysical Models of Large-Scale Multineuronal Data to Enable Precise Neurostimulation arXiv: 2607.04063v1 (2026-07-05) Authors: Amrith Lotlikar, Ian Christopher Tanoh, Praful Vasireddy et al. Categories: q-bio.NC

Problem

Fitting HH biophysical parameters traditionally requires invasive intracellular recordings, which are low-throughput and cannot capture geometry/cell-specific properties of many neurons in a circuit simultaneously. Multi-electrode arrays (MEAs) offer a scalable alternative — high-density extracellular measurements from full neural populations — but HH model complexity has precluded reliable biophysical inference from extracellular data alone.

Solution Architecture

Core Pipeline

Designed MEA Stimulus → Neural Circuit → Extracellular Recordings
                                                    ↓
                              ┌─────────────────────────────────┐
                              │ Differentiable Biophysical       │
                              │ Simulation + Simulation-Based    │
                              │ Inference (SBI)                   │
                              └─────────────────────────────────┘
                                                    ↓
                              HH Parameters (ion channel densities,
                              conductances, morphological params)
                                                    ↓
                              Predict Unseen Stimulation Responses