PINR-ESI: Physics-Informed Self-Supervised Implicit Neural Representation Framework for EEG Extended-Source Imaging
Abstract
EEG extended-source imaging aims to localize active cortical regions and characterize their spatial extent, yet supervised inverse networks typically depend on synthetic EEG-source pairs and may be sensitive to synthetic-to-real mismatch. We propose a physics-informed self-supervised implicit neural representation (INR) framework for spatial source reconstruction without paired source labels or simulation-based supervised pretraining. The lead field matrix encodes the source-to-sensor volume-conduction model and constrains the INR through agreement between predicted and measured EEG. Cortical current density is modeled as a continuous coordinate-conditioned field. A classical inverse estimate provides warm initialization and spatial constraints on source activity, while patch-integrated projection maps the field to discrete source coefficients for forward modeling. Simulations with varying noise levels, source counts, and source extents showed the clearest advantage in identifying active cortical regions, particularly for multiple and larger sources. Quantitative and qualitative evaluations using real EEG recordings from epilepsy patients and intracranial stimulation experiments further supported the feasibility of the framework for source imaging. These findings suggest the potential of physics-informed self-supervised INR for extended-source imaging.
Related
- Neural latent representation of implicit perceptual decision confidence signatures from brain-wide intracranial EEG
- Dual-domain self-supervised feature alignment via spectral–spatial representation learning for deepfake anomaly detection
- PaECTER: Patent-level representation learning using citation-informed transformers
- Semi-supervised clustering with knowledge-guided representation learning in cryo-electron tomography
- AltraFlowSOM: A Semi-Supervised Framework for Imaging Mass Cytometry Phenotyping
BibTeX
@misc{zhou2026pinresi,
title = {{PINR-ESI: Physics-Informed Self-Supervised Implicit Neural Representation Framework for EEG Extended-Source Imaging}},
author = {Zhiyong Zhou and Sheng Liao and Youshen Xiao and Xiaolin Li and Yakang Dai},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.06700},
eprint = {2611.06700},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.06700},
}
RIS
TY - UNPB
TI - PINR-ESI: Physics-Informed Self-Supervised Implicit Neural Representation Framework for EEG Extended-Source Imaging
AU - Zhiyong Zhou
AU - Sheng Liao
AU - Youshen Xiao
AU - Xiaolin Li
AU - Yakang Dai
PY - 2026
DA - 2026/11/04
PB - arXiv
DO - 10.48550/arxiv.2611.06700
UR - https://arxiv.org/abs/2611.06700
AB - EEG extended-source imaging aims to localize active cortical regions and characterize their spatial extent, yet supervised inverse networks typically depend on synthetic EEG-source pairs and may be sensitive to synthetic-to-real mismatch. We propose a physics-informed self-supervised implicit neural representation (INR) framework for spatial source reconstruction without paired source labels or simulation-based supervised pretraining. The lead field matrix encodes the source-to-sensor volume-conduction model and constrains the INR through agreement between predicted and measured EEG. Cortical current density is modeled as a continuous coordinate-conditioned field. A classical inverse estimate provides warm initialization and spatial constraints on source activity, while patch-integrated projection maps the field to discrete source coefficients for forward modeling. Simulations with varying noise levels, source counts, and source extents showed the clearest advantage in identifying active cortical regions, particularly for multiple and larger sources. Quantitative and qualitative evaluations using real EEG recordings from epilepsy patients and intracranial stimulation experiments further supported the feasibility of the framework for source imaging. These findings suggest the potential of physics-informed self-supervised INR for extended-source imaging.
ER -