On the Limits of Applying Graph Transformers for Brain Connectome Classification

Jose Miguel Lara Rangel, Clare Elizabeth Heinbaugh
Proceedings on "I Can't Believe It's Not Better: Challenges in Applied Deep Learning" at ICLR 2025 Workshops, PMLR 296:47-55, 2025.

Abstract

Brain connectomes offer detailed maps of neural connections within the brain. Recent studies have proposed novel connectome graph datasets and attempted to improve connectome classification by using graph deep learning. With recent advances demonstrating transformers’ ability to model intricate relationships and outperform in various domains, this work explores their performance on the novel NeuroGraph benchmark datasets and synthetic variants derived from probabilistically removing edges to simulate noisy data. Our findings suggest that graph transformers offer no major advantage over traditional GNNs on this dataset. Furthermore, both traditional and transformer GNN models maintain accuracy even with all edges removed, suggesting that the dataset’s graph structures may not significantly impact predictions. We propose further assessing NeuroGraph as a brain connectome benchmark, emphasizing the need for well-curated datasets and improved preprocessing strategies to obtain meaningful edge connections.

Cite this Paper

BibTeX
@InProceedings{pmlr-v296-rangel25a, title = {On the Limits of Applying Graph Transformers for Brain Connectome Classification}, author = {Rangel, Jose Miguel Lara and Heinbaugh, Clare Elizabeth}, booktitle = {Proceedings on "I Can't Believe It's Not Better: Challenges in Applied Deep Learning" at ICLR 2025 Workshops}, pages = {47--55}, year = {2025}, editor = {Blaas, Arno and D’Costa, Priya and Feng, Fan and Kriegler, Andreas and Mason, Ian and Pan, Zhaoying and Uelwer, Tobias and Williams, Jennifer and Xie, Yubin and Yang, Rui}, volume = {296}, series = {Proceedings of Machine Learning Research}, month = {28 Apr}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v296/main/assets/rangel25a/rangel25a.pdf}, url = {https://proceedings.mlr.press/v296/rangel25a.html}, abstract = {Brain connectomes offer detailed maps of neural connections within the brain. Recent studies have proposed novel connectome graph datasets and attempted to improve connectome classification by using graph deep learning. With recent advances demonstrating transformers’ ability to model intricate relationships and outperform in various domains, this work explores their performance on the novel NeuroGraph benchmark datasets and synthetic variants derived from probabilistically removing edges to simulate noisy data. Our findings suggest that graph transformers offer no major advantage over traditional GNNs on this dataset. Furthermore, both traditional and transformer GNN models maintain accuracy even with all edges removed, suggesting that the dataset’s graph structures may not significantly impact predictions. We propose further assessing NeuroGraph as a brain connectome benchmark, emphasizing the need for well-curated datasets and improved preprocessing strategies to obtain meaningful edge connections.} }
Endnote
%0 Conference Paper %T On the Limits of Applying Graph Transformers for Brain Connectome Classification %A Jose Miguel Lara Rangel %A Clare Elizabeth Heinbaugh %B Proceedings on "I Can't Believe It's Not Better: Challenges in Applied Deep Learning" at ICLR 2025 Workshops %C Proceedings of Machine Learning Research %D 2025 %E Arno Blaas %E Priya D’Costa %E Fan Feng %E Andreas Kriegler %E Ian Mason %E Zhaoying Pan %E Tobias Uelwer %E Jennifer Williams %E Yubin Xie %E Rui Yang %F pmlr-v296-rangel25a %I PMLR %P 47--55 %U https://proceedings.mlr.press/v296/rangel25a.html %V 296 %X Brain connectomes offer detailed maps of neural connections within the brain. Recent studies have proposed novel connectome graph datasets and attempted to improve connectome classification by using graph deep learning. With recent advances demonstrating transformers’ ability to model intricate relationships and outperform in various domains, this work explores their performance on the novel NeuroGraph benchmark datasets and synthetic variants derived from probabilistically removing edges to simulate noisy data. Our findings suggest that graph transformers offer no major advantage over traditional GNNs on this dataset. Furthermore, both traditional and transformer GNN models maintain accuracy even with all edges removed, suggesting that the dataset’s graph structures may not significantly impact predictions. We propose further assessing NeuroGraph as a brain connectome benchmark, emphasizing the need for well-curated datasets and improved preprocessing strategies to obtain meaningful edge connections.
APA
Rangel, J.M.L. & Heinbaugh, C.E.. (2025). On the Limits of Applying Graph Transformers for Brain Connectome Classification. Proceedings on "I Can't Believe It's Not Better: Challenges in Applied Deep Learning" at ICLR 2025 Workshops, in Proceedings of Machine Learning Research 296:47-55 Available from https://proceedings.mlr.press/v296/rangel25a.html.

Related Material