Temporal PHATE: A multi-view manifold learning method for brain state trajectories. Busch, E. L., Huang, J., Benz, A., Wallenstein, T., Lajoie, G., Wolf, G., Krishnaswamy, S., & Turk-Browne, N. B. Technical Report bioRxiv, May, 2022. Section: New Results Type: article![Temporal PHATE: A multi-view manifold learning method for brain state trajectories [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex Brain activity as measured with functional magnetic resonance imaging (fMRI) gives the illusion of intractably high dimensionality, rife with collection and biological noise. Nonlinear dimensionality reductions like UMAP, tSNE, and PHATE have proven useful for high-throughput biomedical data, but have not been extensively used in fMRI, which is known to reflect the redundancy and co-modulation of neural population activity. Here we take the manifold-geometry preserving method PHATE and extend it for use in brain activity timeseries data in a method we call temporal PHATE (T-PHATE). We observe that in addition to the intrinsically lower dimensionality of fMRI data, it also has significant autocorrelative structure that we can exploit to faithfully denoise the signal and learn brain activation manifolds. We empirically validate T-PHATE on three fMRI tasks and show that T-PHATE manifolds improve visualization fidelity, stimulus feature classification, and neural event segmentation. T-PHATE demonstrates impressive improvements over previous cutting-edge approaches to understanding the nature of cognition from fMRI and bodes potential applications broadly for high-dimensional datasets of temporally-diffuse processes.
@techreport{busch_temporal_2022,
title = {Temporal {PHATE}: {A} multi-view manifold learning method for brain state trajectories},
copyright = {© 2022, Posted by Cold Spring Harbor Laboratory. This pre-print is available under a Creative Commons License (Attribution-NonCommercial-NoDerivs 4.0 International), CC BY-NC-ND 4.0, as described at http://creativecommons.org/licenses/by-nc-nd/4.0/},
shorttitle = {Temporal {PHATE}},
url = {https://www.biorxiv.org/content/10.1101/2022.05.03.490534v2},
abstract = {Brain activity as measured with functional magnetic resonance imaging (fMRI) gives the illusion of intractably high dimensionality, rife with collection and biological noise. Nonlinear dimensionality reductions like UMAP, tSNE, and PHATE have proven useful for high-throughput biomedical data, but have not been extensively used in fMRI, which is known to reflect the redundancy and co-modulation of neural population activity. Here we take the manifold-geometry preserving method PHATE and extend it for use in brain activity timeseries data in a method we call temporal PHATE (T-PHATE). We observe that in addition to the intrinsically lower dimensionality of fMRI data, it also has significant autocorrelative structure that we can exploit to faithfully denoise the signal and learn brain activation manifolds. We empirically validate T-PHATE on three fMRI tasks and show that T-PHATE manifolds improve visualization fidelity, stimulus feature classification, and neural event segmentation. T-PHATE demonstrates impressive improvements over previous cutting-edge approaches to understanding the nature of cognition from fMRI and bodes potential applications broadly for high-dimensional datasets of temporally-diffuse processes.},
language = {en},
urldate = {2022-05-06},
institution = {bioRxiv},
author = {Busch, Erica L. and Huang, Jessie and Benz, Andrew and Wallenstein, Tom and Lajoie, Guillaume and Wolf, Guy and Krishnaswamy, Smita and Turk-Browne, Nicholas B.},
month = may,
year = {2022},
doi = {10.1101/2022.05.03.490534},
note = {Section: New Results
Type: article},
pages = {2022.05.03.490534},
}
Downloads: 0
{"_id":"vTCLyEgf8wDkx2jXc","bibbaseid":"busch-huang-benz-wallenstein-lajoie-wolf-krishnaswamy-turkbrowne-temporalphateamultiviewmanifoldlearningmethodforbrainstatetrajectories-2022","author_short":["Busch, E. L.","Huang, J.","Benz, A.","Wallenstein, T.","Lajoie, G.","Wolf, G.","Krishnaswamy, S.","Turk-Browne, N. B."],"bibdata":{"bibtype":"techreport","type":"techreport","title":"Temporal PHATE: A multi-view manifold learning method for brain state trajectories","copyright":"© 2022, Posted by Cold Spring Harbor Laboratory. This pre-print is available under a Creative Commons License (Attribution-NonCommercial-NoDerivs 4.0 International), CC BY-NC-ND 4.0, as described at http://creativecommons.org/licenses/by-nc-nd/4.0/","shorttitle":"Temporal PHATE","url":"https://www.biorxiv.org/content/10.1101/2022.05.03.490534v2","abstract":"Brain activity as measured with functional magnetic resonance imaging (fMRI) gives the illusion of intractably high dimensionality, rife with collection and biological noise. Nonlinear dimensionality reductions like UMAP, tSNE, and PHATE have proven useful for high-throughput biomedical data, but have not been extensively used in fMRI, which is known to reflect the redundancy and co-modulation of neural population activity. Here we take the manifold-geometry preserving method PHATE and extend it for use in brain activity timeseries data in a method we call temporal PHATE (T-PHATE). We observe that in addition to the intrinsically lower dimensionality of fMRI data, it also has significant autocorrelative structure that we can exploit to faithfully denoise the signal and learn brain activation manifolds. We empirically validate T-PHATE on three fMRI tasks and show that T-PHATE manifolds improve visualization fidelity, stimulus feature classification, and neural event segmentation. T-PHATE demonstrates impressive improvements over previous cutting-edge approaches to understanding the nature of cognition from fMRI and bodes potential applications broadly for high-dimensional datasets of temporally-diffuse processes.","language":"en","urldate":"2022-05-06","institution":"bioRxiv","author":[{"propositions":[],"lastnames":["Busch"],"firstnames":["Erica","L."],"suffixes":[]},{"propositions":[],"lastnames":["Huang"],"firstnames":["Jessie"],"suffixes":[]},{"propositions":[],"lastnames":["Benz"],"firstnames":["Andrew"],"suffixes":[]},{"propositions":[],"lastnames":["Wallenstein"],"firstnames":["Tom"],"suffixes":[]},{"propositions":[],"lastnames":["Lajoie"],"firstnames":["Guillaume"],"suffixes":[]},{"propositions":[],"lastnames":["Wolf"],"firstnames":["Guy"],"suffixes":[]},{"propositions":[],"lastnames":["Krishnaswamy"],"firstnames":["Smita"],"suffixes":[]},{"propositions":[],"lastnames":["Turk-Browne"],"firstnames":["Nicholas","B."],"suffixes":[]}],"month":"May","year":"2022","doi":"10.1101/2022.05.03.490534","note":"Section: New Results Type: article","pages":"2022.05.03.490534","bibtex":"@techreport{busch_temporal_2022,\n\ttitle = {Temporal {PHATE}: {A} multi-view manifold learning method for brain state trajectories},\n\tcopyright = {© 2022, Posted by Cold Spring Harbor Laboratory. This pre-print is available under a Creative Commons License (Attribution-NonCommercial-NoDerivs 4.0 International), CC BY-NC-ND 4.0, as described at http://creativecommons.org/licenses/by-nc-nd/4.0/},\n\tshorttitle = {Temporal {PHATE}},\n\turl = {https://www.biorxiv.org/content/10.1101/2022.05.03.490534v2},\n\tabstract = {Brain activity as measured with functional magnetic resonance imaging (fMRI) gives the illusion of intractably high dimensionality, rife with collection and biological noise. Nonlinear dimensionality reductions like UMAP, tSNE, and PHATE have proven useful for high-throughput biomedical data, but have not been extensively used in fMRI, which is known to reflect the redundancy and co-modulation of neural population activity. Here we take the manifold-geometry preserving method PHATE and extend it for use in brain activity timeseries data in a method we call temporal PHATE (T-PHATE). We observe that in addition to the intrinsically lower dimensionality of fMRI data, it also has significant autocorrelative structure that we can exploit to faithfully denoise the signal and learn brain activation manifolds. We empirically validate T-PHATE on three fMRI tasks and show that T-PHATE manifolds improve visualization fidelity, stimulus feature classification, and neural event segmentation. T-PHATE demonstrates impressive improvements over previous cutting-edge approaches to understanding the nature of cognition from fMRI and bodes potential applications broadly for high-dimensional datasets of temporally-diffuse processes.},\n\tlanguage = {en},\n\turldate = {2022-05-06},\n\tinstitution = {bioRxiv},\n\tauthor = {Busch, Erica L. and Huang, Jessie and Benz, Andrew and Wallenstein, Tom and Lajoie, Guillaume and Wolf, Guy and Krishnaswamy, Smita and Turk-Browne, Nicholas B.},\n\tmonth = may,\n\tyear = {2022},\n\tdoi = {10.1101/2022.05.03.490534},\n\tnote = {Section: New Results\nType: article},\n\tpages = {2022.05.03.490534},\n}\n\n","author_short":["Busch, E. L.","Huang, J.","Benz, A.","Wallenstein, T.","Lajoie, G.","Wolf, G.","Krishnaswamy, S.","Turk-Browne, N. B."],"key":"busch_temporal_2022","id":"busch_temporal_2022","bibbaseid":"busch-huang-benz-wallenstein-lajoie-wolf-krishnaswamy-turkbrowne-temporalphateamultiviewmanifoldlearningmethodforbrainstatetrajectories-2022","role":"author","urls":{"Paper":"https://www.biorxiv.org/content/10.1101/2022.05.03.490534v2"},"metadata":{"authorlinks":{}}},"bibtype":"techreport","biburl":"https://api.zotero.org/groups/4553800/items?key=BfP7bN7FF9dJwtyiLBORewdg&format=bibtex&limit=100","dataSources":["5EcJ4DjTLGyJLTga3"],"keywords":[],"search_terms":["temporal","phate","multi","view","manifold","learning","method","brain","state","trajectories","busch","huang","benz","wallenstein","lajoie","wolf","krishnaswamy","turk-browne"],"title":"Temporal PHATE: A multi-view manifold learning method for brain state trajectories","year":2022}