Learning Cortical Parcellations Using Graph Neural Networks
Deep learning has been applied to magnetic resonance imaging (MRI) for a variety of purposes, ranging from the acceleration of image acquisition and image denoising to tissue segmentation and disease diagnosis. Convolutional neural networks have been particularly useful for analyzing MRI data due to...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2021-12-01
|
| Series: | Frontiers in Neuroscience |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fnins.2021.797500/full |
| _version_ | 1819096018968379392 |
|---|---|
| author | Kristian M. Eschenburg Kristian M. Eschenburg Thomas J. Grabowski Thomas J. Grabowski Thomas J. Grabowski David R. Haynor David R. Haynor David R. Haynor |
| author_facet | Kristian M. Eschenburg Kristian M. Eschenburg Thomas J. Grabowski Thomas J. Grabowski Thomas J. Grabowski David R. Haynor David R. Haynor David R. Haynor |
| author_sort | Kristian M. Eschenburg |
| collection | DOAJ |
| description | Deep learning has been applied to magnetic resonance imaging (MRI) for a variety of purposes, ranging from the acceleration of image acquisition and image denoising to tissue segmentation and disease diagnosis. Convolutional neural networks have been particularly useful for analyzing MRI data due to the regularly sampled spatial and temporal nature of the data. However, advances in the field of brain imaging have led to network- and surface-based analyses that are often better represented in the graph domain. In this analysis, we propose a general purpose cortical segmentation method that, given resting-state connectivity features readily computed during conventional MRI pre-processing and a set of corresponding training labels, can generate cortical parcellations for new MRI data. We applied recent advances in the field of graph neural networks to the problem of cortical surface segmentation, using resting-state connectivity to learn discrete maps of the human neocortex. We found that graph neural networks accurately learn low-dimensional representations of functional brain connectivity that can be naturally extended to map the cortices of new datasets. After optimizing over algorithm type, network architecture, and training features, our approach yielded mean classification accuracies of 79.91% relative to a previously published parcellation. We describe how some hyperparameter choices including training and testing data duration, network architecture, and algorithm choice affect model performance. |
| first_indexed | 2024-12-21T23:52:32Z |
| format | Article |
| id | doaj.art-cf268ba1e0b4487485303ef546ed0894 |
| institution | Directory Open Access Journal |
| issn | 1662-453X |
| language | English |
| last_indexed | 2024-12-21T23:52:32Z |
| publishDate | 2021-12-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neuroscience |
| spelling | doaj.art-cf268ba1e0b4487485303ef546ed08942022-12-21T18:45:54ZengFrontiers Media S.A.Frontiers in Neuroscience1662-453X2021-12-011510.3389/fnins.2021.797500797500Learning Cortical Parcellations Using Graph Neural NetworksKristian M. Eschenburg0Kristian M. Eschenburg1Thomas J. Grabowski2Thomas J. Grabowski3Thomas J. Grabowski4David R. Haynor5David R. Haynor6David R. Haynor7Department of Bioengineering, University of Washington, Seattle, WA, United StatesIntegrated Brain Imaging Center, University of Washington Medical Center, Seattle, WA, United StatesIntegrated Brain Imaging Center, University of Washington Medical Center, Seattle, WA, United StatesDepartment of Radiology, University of Washington Medical Center, Seattle, WA, United StatesDepartment of Neurology, University of Washington Medical Center, Seattle, WA, United StatesDepartment of Bioengineering, University of Washington, Seattle, WA, United StatesIntegrated Brain Imaging Center, University of Washington Medical Center, Seattle, WA, United StatesDepartment of Radiology, University of Washington Medical Center, Seattle, WA, United StatesDeep learning has been applied to magnetic resonance imaging (MRI) for a variety of purposes, ranging from the acceleration of image acquisition and image denoising to tissue segmentation and disease diagnosis. Convolutional neural networks have been particularly useful for analyzing MRI data due to the regularly sampled spatial and temporal nature of the data. However, advances in the field of brain imaging have led to network- and surface-based analyses that are often better represented in the graph domain. In this analysis, we propose a general purpose cortical segmentation method that, given resting-state connectivity features readily computed during conventional MRI pre-processing and a set of corresponding training labels, can generate cortical parcellations for new MRI data. We applied recent advances in the field of graph neural networks to the problem of cortical surface segmentation, using resting-state connectivity to learn discrete maps of the human neocortex. We found that graph neural networks accurately learn low-dimensional representations of functional brain connectivity that can be naturally extended to map the cortices of new datasets. After optimizing over algorithm type, network architecture, and training features, our approach yielded mean classification accuracies of 79.91% relative to a previously published parcellation. We describe how some hyperparameter choices including training and testing data duration, network architecture, and algorithm choice affect model performance.https://www.frontiersin.org/articles/10.3389/fnins.2021.797500/fullgraph neural networkparcellationfunctional connectivityrepresentation learningsegmentationbrain |
| spellingShingle | Kristian M. Eschenburg Kristian M. Eschenburg Thomas J. Grabowski Thomas J. Grabowski Thomas J. Grabowski David R. Haynor David R. Haynor David R. Haynor Learning Cortical Parcellations Using Graph Neural Networks Frontiers in Neuroscience graph neural network parcellation functional connectivity representation learning segmentation brain |
| title | Learning Cortical Parcellations Using Graph Neural Networks |
| title_full | Learning Cortical Parcellations Using Graph Neural Networks |
| title_fullStr | Learning Cortical Parcellations Using Graph Neural Networks |
| title_full_unstemmed | Learning Cortical Parcellations Using Graph Neural Networks |
| title_short | Learning Cortical Parcellations Using Graph Neural Networks |
| title_sort | learning cortical parcellations using graph neural networks |
| topic | graph neural network parcellation functional connectivity representation learning segmentation brain |
| url | https://www.frontiersin.org/articles/10.3389/fnins.2021.797500/full |
| work_keys_str_mv | AT kristianmeschenburg learningcorticalparcellationsusinggraphneuralnetworks AT kristianmeschenburg learningcorticalparcellationsusinggraphneuralnetworks AT thomasjgrabowski learningcorticalparcellationsusinggraphneuralnetworks AT thomasjgrabowski learningcorticalparcellationsusinggraphneuralnetworks AT thomasjgrabowski learningcorticalparcellationsusinggraphneuralnetworks AT davidrhaynor learningcorticalparcellationsusinggraphneuralnetworks AT davidrhaynor learningcorticalparcellationsusinggraphneuralnetworks AT davidrhaynor learningcorticalparcellationsusinggraphneuralnetworks |