Speaker
Description
Biological neural networks (BNNs) are machine learning models that enhance the biological interpretability of artificial neural networks by modeling neural dynamics and providing insights underlying neural system behavior. In our previous work, Hodgkin-Huxley neuron models were implemented in BNNs to classify electroencephalogram (EEG) signals \cite{cruz2026bnn}. While biologically interpretable, these models treat each electrode independently and ignore spatial relationships, limiting their ability to capture large-scale brain dynamics.
Here, we extend this framework by developing trainable BNNs using time-aware backpropagation applied to networks of biophysically accurate neurons based on modified Hodgkin-Huxley equations \cite{deistler2024diffsim}, integrated within feedforward and recurrent architectures. Neuron sets represent brain regions, and synaptic weights reflect spatial distances and functional connectivity. These BNNs classify consciousness levels from EEG data, extrapolate deeper brain activity, and exhibit physiologically observed frequencies, while hidden-layer neurons remain biologically interpretable.
Overall, spatially informed BNNs capture deeper structures, generate emergent spatiotemporal patterns, and improve cross-subject generalization, advancing both EEG interpretability and generalizability while bridging neural modeling and modern machine learning.
Bibliography
@misc{cruz2026bnn,
author = {Cruz, Madelyn Esther Chua and Hazelden, James and Ivanitskiy, Michael I. and D'Alessandro, Matthew and Negelspach, David Casey and Huskey, Alisa and Killgore, William D. S. and Mashour, George A. and Forger, Daniel B.},
title = {Biologically Interpretable Machine Learning Approaches for Analyzing Neural Data},
note = {Submitted for publication},
year = {2026}
}
@article{deistler2024diffsim,
author = {Deistler, Maximilian and Kadhim, K. L. and Pals, M. and Beck, J. and Huang, Z. and Gloeckler, M. and ... and Macke, J. H.},
title = {Differentiable simulation enables large-scale training of detailed biophysical models of neural dynamics},
journal = {bioRxiv},
year = {2024},
month = {August},
note = {2024-08}
}
