Evaluating topological fitness of human brain-inspired sub-circuits in Echo State Networks

Bach V. Nguyen, Tianlong Chen, Shu Yang, Bojian Hou, Li Shen, Duy Duong-Tran
Proceedings of The First AAAI Bridge Program on AI for Medicine and Healthcare, PMLR 281:79-89, 2025.

Abstract

In recent years, an emerging trend in neuromorphic computing has centered around the use of brain connectomics as a blueprint for artificial neural networks. Connectomics-based neuromorphic computing has primarily focused on embedding human brain large-scale structural connectomes (SCs), as estimated from diffusion Magnetic Resonance Imaging (dMRI) modality, to echo-state networks (ESNs). A critical step in ESN embedding requires pre-determined read-in and read-out layers constructed by the induced subgraphs of the embedded reservoir. As a priori set of functional sub-circuits are derived from functional MRI (fMRI) modality, it is unknown, till this point, whether the embedding of fMRI-induced sub-circuits/networks onto SCs is well justified from the neuro-physiological perspective and ESN performance across a variety of tasks. This paper proposes a pipeline to implement and evaluate ESNs with various embedded topologies and processing/memorization tasks. To this end, we showed that different performance optimums highly depend on the neuro-physiological characteristics of these pre-determined fMRI-induced sub-circuits. In general, fMRI-induced sub-circuit-embedded ESN outperforms simple bipartite and various null models with feed-forward properties commonly seen in MLP for different tasks and reservoir criticality conditions. We provided a thorough analysis of the topological properties of pre-determined fMRI-induced sub-circuits and highlighted their graph-theoretical properties that play significant roles in determining ESN performance. Finally, we demonstrate the model’s performance in predicting epidemiological time-series COVID-19 datasets, showing the bio-inspired model’s potential in application to public health decision-making.

Cite this Paper

BibTeX
@InProceedings{pmlr-v281-nguyen25a, title = {Evaluating topological fitness of human brain-inspired sub-circuits in Echo State Networks}, author = {Nguyen, Bach V. and Chen, Tianlong and Yang, Shu and Hou, Bojian and Shen, Li and Duong-Tran, Duy}, booktitle = {Proceedings of The First AAAI Bridge Program on AI for Medicine and Healthcare}, pages = {79--89}, year = {2025}, editor = {Wu, Junde and Zhu, Jiayuan and Xu, Min and Jin, Yueming}, volume = {281}, series = {Proceedings of Machine Learning Research}, month = {25 Feb}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v281/main/assets/nguyen25a/nguyen25a.pdf}, url = {https://proceedings.mlr.press/v281/nguyen25a.html}, abstract = {In recent years, an emerging trend in neuromorphic computing has centered around the use of brain connectomics as a blueprint for artificial neural networks. Connectomics-based neuromorphic computing has primarily focused on embedding human brain large-scale structural connectomes (SCs), as estimated from diffusion Magnetic Resonance Imaging (dMRI) modality, to echo-state networks (ESNs). A critical step in ESN embedding requires pre-determined read-in and read-out layers constructed by the induced subgraphs of the embedded reservoir. As a priori set of functional sub-circuits are derived from functional MRI (fMRI) modality, it is unknown, till this point, whether the embedding of fMRI-induced sub-circuits/networks onto SCs is well justified from the neuro-physiological perspective and ESN performance across a variety of tasks. This paper proposes a pipeline to implement and evaluate ESNs with various embedded topologies and processing/memorization tasks. To this end, we showed that different performance optimums highly depend on the neuro-physiological characteristics of these pre-determined fMRI-induced sub-circuits. In general, fMRI-induced sub-circuit-embedded ESN outperforms simple bipartite and various null models with feed-forward properties commonly seen in MLP for different tasks and reservoir criticality conditions. We provided a thorough analysis of the topological properties of pre-determined fMRI-induced sub-circuits and highlighted their graph-theoretical properties that play significant roles in determining ESN performance. Finally, we demonstrate the model’s performance in predicting epidemiological time-series COVID-19 datasets, showing the bio-inspired model’s potential in application to public health decision-making.} }
Endnote
%0 Conference Paper %T Evaluating topological fitness of human brain-inspired sub-circuits in Echo State Networks %A Bach V. Nguyen %A Tianlong Chen %A Shu Yang %A Bojian Hou %A Li Shen %A Duy Duong-Tran %B Proceedings of The First AAAI Bridge Program on AI for Medicine and Healthcare %C Proceedings of Machine Learning Research %D 2025 %E Junde Wu %E Jiayuan Zhu %E Min Xu %E Yueming Jin %F pmlr-v281-nguyen25a %I PMLR %P 79--89 %U https://proceedings.mlr.press/v281/nguyen25a.html %V 281 %X In recent years, an emerging trend in neuromorphic computing has centered around the use of brain connectomics as a blueprint for artificial neural networks. Connectomics-based neuromorphic computing has primarily focused on embedding human brain large-scale structural connectomes (SCs), as estimated from diffusion Magnetic Resonance Imaging (dMRI) modality, to echo-state networks (ESNs). A critical step in ESN embedding requires pre-determined read-in and read-out layers constructed by the induced subgraphs of the embedded reservoir. As a priori set of functional sub-circuits are derived from functional MRI (fMRI) modality, it is unknown, till this point, whether the embedding of fMRI-induced sub-circuits/networks onto SCs is well justified from the neuro-physiological perspective and ESN performance across a variety of tasks. This paper proposes a pipeline to implement and evaluate ESNs with various embedded topologies and processing/memorization tasks. To this end, we showed that different performance optimums highly depend on the neuro-physiological characteristics of these pre-determined fMRI-induced sub-circuits. In general, fMRI-induced sub-circuit-embedded ESN outperforms simple bipartite and various null models with feed-forward properties commonly seen in MLP for different tasks and reservoir criticality conditions. We provided a thorough analysis of the topological properties of pre-determined fMRI-induced sub-circuits and highlighted their graph-theoretical properties that play significant roles in determining ESN performance. Finally, we demonstrate the model’s performance in predicting epidemiological time-series COVID-19 datasets, showing the bio-inspired model’s potential in application to public health decision-making.
APA
Nguyen, B.V., Chen, T., Yang, S., Hou, B., Shen, L. & Duong-Tran, D.. (2025). Evaluating topological fitness of human brain-inspired sub-circuits in Echo State Networks. Proceedings of The First AAAI Bridge Program on AI for Medicine and Healthcare, in Proceedings of Machine Learning Research 281:79-89 Available from https://proceedings.mlr.press/v281/nguyen25a.html.

Related Material