Hybrid Dynamic High-Order Functional Correlations and Divisive Normalization for Improved Classification of Schizophrenia and Bipolar Disorder

Qiang Li, Vince Calhoun
Proceedings of UniReps: the Second Edition of the Workshop on Unifying Representations in Neural Models, PMLR 285:170-180, 2024.

Abstract

Schizophrenia and bipolar disorder are devastating psychiatric disorders that can be difficult to adequately classify, considering commonalities that make it difficult to distinguish between them using conventional classification approaches based on low-order functional connectivity. Recently, high-order functional connectivity has emerged as a promising method for diagnosing psychiatric illnesses, and this research applied multiple strategies for distinguishing schizophrenia and bipolar disorder using features taken from dynamic high-order functional connectivity and divisive normalization. The approach that produced the greatest results combined dynamic high-order functional correlations and divisive normalization to examine patterns of intrinsic connection time courses collected from resting-state fMRI. Our findings indicate that resting-state fMRI-based dynamic high-order functional connectivity and feature enhancement through divisive normalization classification hold significant promise for improving the accuracy of psychiatric diagnoses. Moreover, to the best of our knowledge, this study is the first to integrate divisive normalization with functional connectivity in fMRI.

Cite this Paper

BibTeX
@InProceedings{pmlr-v285-li24a, title = {Hybrid Dynamic High-Order Functional Correlations and Divisive Normalization for Improved Classification of Schizophrenia and Bipolar Disorder}, author = {Li, Qiang and Calhoun, Vince}, booktitle = {Proceedings of UniReps: the Second Edition of the Workshop on Unifying Representations in Neural Models}, pages = {170--180}, year = {2024}, editor = {Fumero, Marco and Domine, Clementine and Lähner, Zorah and Crisostomi, Donato and Moschella, Luca and Stachenfeld, Kimberly}, volume = {285}, series = {Proceedings of Machine Learning Research}, month = {14 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v285/main/assets/li24a/li24a.pdf}, url = {https://proceedings.mlr.press/v285/li24a.html}, abstract = {Schizophrenia and bipolar disorder are devastating psychiatric disorders that can be difficult to adequately classify, considering commonalities that make it difficult to distinguish between them using conventional classification approaches based on low-order functional connectivity. Recently, high-order functional connectivity has emerged as a promising method for diagnosing psychiatric illnesses, and this research applied multiple strategies for distinguishing schizophrenia and bipolar disorder using features taken from dynamic high-order functional connectivity and divisive normalization. The approach that produced the greatest results combined dynamic high-order functional correlations and divisive normalization to examine patterns of intrinsic connection time courses collected from resting-state fMRI. Our findings indicate that resting-state fMRI-based dynamic high-order functional connectivity and feature enhancement through divisive normalization classification hold significant promise for improving the accuracy of psychiatric diagnoses. Moreover, to the best of our knowledge, this study is the first to integrate divisive normalization with functional connectivity in fMRI.} }
Endnote
%0 Conference Paper %T Hybrid Dynamic High-Order Functional Correlations and Divisive Normalization for Improved Classification of Schizophrenia and Bipolar Disorder %A Qiang Li %A Vince Calhoun %B Proceedings of UniReps: the Second Edition of the Workshop on Unifying Representations in Neural Models %C Proceedings of Machine Learning Research %D 2024 %E Marco Fumero %E Clementine Domine %E Zorah Lähner %E Donato Crisostomi %E Luca Moschella %E Kimberly Stachenfeld %F pmlr-v285-li24a %I PMLR %P 170--180 %U https://proceedings.mlr.press/v285/li24a.html %V 285 %X Schizophrenia and bipolar disorder are devastating psychiatric disorders that can be difficult to adequately classify, considering commonalities that make it difficult to distinguish between them using conventional classification approaches based on low-order functional connectivity. Recently, high-order functional connectivity has emerged as a promising method for diagnosing psychiatric illnesses, and this research applied multiple strategies for distinguishing schizophrenia and bipolar disorder using features taken from dynamic high-order functional connectivity and divisive normalization. The approach that produced the greatest results combined dynamic high-order functional correlations and divisive normalization to examine patterns of intrinsic connection time courses collected from resting-state fMRI. Our findings indicate that resting-state fMRI-based dynamic high-order functional connectivity and feature enhancement through divisive normalization classification hold significant promise for improving the accuracy of psychiatric diagnoses. Moreover, to the best of our knowledge, this study is the first to integrate divisive normalization with functional connectivity in fMRI.
APA
Li, Q. & Calhoun, V.. (2024). Hybrid Dynamic High-Order Functional Correlations and Divisive Normalization for Improved Classification of Schizophrenia and Bipolar Disorder. Proceedings of UniReps: the Second Edition of the Workshop on Unifying Representations in Neural Models, in Proceedings of Machine Learning Research 285:170-180 Available from https://proceedings.mlr.press/v285/li24a.html.

Related Material