Cargando…

Stacked regressions and structured variance partitioning for interpretable brain maps

Relating brain activity associated with a complex stimulus to different properties of that stimulus is a powerful approach for constructing functional brain maps. However, when stimuli are naturalistic, their properties are often correlated (e.g., visual and semantic features of natural images, or d...

Descripción completa

Detalles Bibliográficos
Autores principales: Lin, Ruogu, Naselaris, Thomas, Kay, Kendrick, Wehbe, Leila
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Cold Spring Harbor Laboratory 2023
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10168225/
https://www.ncbi.nlm.nih.gov/pubmed/37163111
http://dx.doi.org/10.1101/2023.04.23.537988
_version_ 1785038819181985792
author Lin, Ruogu
Naselaris, Thomas
Kay, Kendrick
Wehbe, Leila
author_facet Lin, Ruogu
Naselaris, Thomas
Kay, Kendrick
Wehbe, Leila
author_sort Lin, Ruogu
collection PubMed
description Relating brain activity associated with a complex stimulus to different properties of that stimulus is a powerful approach for constructing functional brain maps. However, when stimuli are naturalistic, their properties are often correlated (e.g., visual and semantic features of natural images, or different layers of a convolutional neural network that are used as features of images). Correlated properties can act as confounders for each other and complicate the interpretability of brain maps, and can impact the robustness of statistical estimators. Here, we present an approach for brain mapping based on two proposed methods: stacking different encoding models and structured variance partitioning. Our stacking algorithm combines encoding models that each use as input a feature space that describes a different stimulus attribute. The algorithm learns to predict the activity of a voxel as a linear combination of the outputs of different encoding models. We show that the resulting combined model can predict held-out brain activity better or at least as well as the individual encoding models. Further, the weights of the linear combination are readily interpretable; they show the importance of each feature space for predicting a voxel. We then build on our stacking models to introduce structured variance partitioning, a new type of variance partitioning that takes into account the known relationships between features. Our approach constrains the size of the hypothesis space and allows us to ask targeted questions about the similarity between feature spaces and brain regions even in the presence of correlations between the feature spaces. We validate our approach in simulation, showcase its brain mapping potential on fMRI data, and release a Python package. Our methods can be useful for researchers interested in aligning brain activity with different layers of a neural network, or with other types of correlated feature spaces.
format Online
Article
Text
id pubmed-10168225
institution National Center for Biotechnology Information
language English
publishDate 2023
publisher Cold Spring Harbor Laboratory
record_format MEDLINE/PubMed
spelling pubmed-101682252023-05-10 Stacked regressions and structured variance partitioning for interpretable brain maps Lin, Ruogu Naselaris, Thomas Kay, Kendrick Wehbe, Leila bioRxiv Article Relating brain activity associated with a complex stimulus to different properties of that stimulus is a powerful approach for constructing functional brain maps. However, when stimuli are naturalistic, their properties are often correlated (e.g., visual and semantic features of natural images, or different layers of a convolutional neural network that are used as features of images). Correlated properties can act as confounders for each other and complicate the interpretability of brain maps, and can impact the robustness of statistical estimators. Here, we present an approach for brain mapping based on two proposed methods: stacking different encoding models and structured variance partitioning. Our stacking algorithm combines encoding models that each use as input a feature space that describes a different stimulus attribute. The algorithm learns to predict the activity of a voxel as a linear combination of the outputs of different encoding models. We show that the resulting combined model can predict held-out brain activity better or at least as well as the individual encoding models. Further, the weights of the linear combination are readily interpretable; they show the importance of each feature space for predicting a voxel. We then build on our stacking models to introduce structured variance partitioning, a new type of variance partitioning that takes into account the known relationships between features. Our approach constrains the size of the hypothesis space and allows us to ask targeted questions about the similarity between feature spaces and brain regions even in the presence of correlations between the feature spaces. We validate our approach in simulation, showcase its brain mapping potential on fMRI data, and release a Python package. Our methods can be useful for researchers interested in aligning brain activity with different layers of a neural network, or with other types of correlated feature spaces. Cold Spring Harbor Laboratory 2023-04-24 /pmc/articles/PMC10168225/ /pubmed/37163111 http://dx.doi.org/10.1101/2023.04.23.537988 Text en https://creativecommons.org/licenses/by-nd/4.0/This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License (https://creativecommons.org/licenses/by-nd/4.0/) , which allows reusers to copy and distribute the material in any medium or format in unadapted form only, and only so long as attribution is given to the creator. The license allows for commercial use.
spellingShingle Article
Lin, Ruogu
Naselaris, Thomas
Kay, Kendrick
Wehbe, Leila
Stacked regressions and structured variance partitioning for interpretable brain maps
title Stacked regressions and structured variance partitioning for interpretable brain maps
title_full Stacked regressions and structured variance partitioning for interpretable brain maps
title_fullStr Stacked regressions and structured variance partitioning for interpretable brain maps
title_full_unstemmed Stacked regressions and structured variance partitioning for interpretable brain maps
title_short Stacked regressions and structured variance partitioning for interpretable brain maps
title_sort stacked regressions and structured variance partitioning for interpretable brain maps
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10168225/
https://www.ncbi.nlm.nih.gov/pubmed/37163111
http://dx.doi.org/10.1101/2023.04.23.537988
work_keys_str_mv AT linruogu stackedregressionsandstructuredvariancepartitioningforinterpretablebrainmaps
AT naselaristhomas stackedregressionsandstructuredvariancepartitioningforinterpretablebrainmaps
AT kaykendrick stackedregressionsandstructuredvariancepartitioningforinterpretablebrainmaps
AT wehbeleila stackedregressionsandstructuredvariancepartitioningforinterpretablebrainmaps