TY - GEN
T1 - A novel graph neural network to localize eloquent cortex in brain tumor patients from resting-state FMRI connectivity
AU - Nandakumar, Naresh
AU - Manzoor, Komal
AU - Pillai, Jay J.
AU - Gujar, Sachin K.
AU - Sair, Haris I.
AU - Venkataraman, Archana
N1 - Publisher Copyright:
© Springer Nature Switzerland AG 2019.
PY - 2019
Y1 - 2019
N2 - We develop a novel method to localize the language and motor areas of the eloquent cortex in brain tumor patients based on resting-state fMRI (rs-fMRI) connectivity. Our method leverages the representation power of convolutional neural networks through specialized filters that act topologically on the rs-fMRI connectivity data. This Graph Neural Network (GNN) classifies each parcel in the brain into eloquent cortex, tumor, or background gray matter, thus accommodating varying tumor characteristics across patients. Our loss function also reflects the large class-imbalance present in our data. We evaluate our GNN on rs-fMRI data from 60 brain tumor patients with different tumor sizes and locations. We use motor and language task fMRI for validation. Our model achieves better localization than linear SVM, random forest, and a multilayer perceptron architecture. Our GNN is able to correctly identify bilateral language areas in the brain even when trained on patients whose language network is lateralized to the left hemisphere.
AB - We develop a novel method to localize the language and motor areas of the eloquent cortex in brain tumor patients based on resting-state fMRI (rs-fMRI) connectivity. Our method leverages the representation power of convolutional neural networks through specialized filters that act topologically on the rs-fMRI connectivity data. This Graph Neural Network (GNN) classifies each parcel in the brain into eloquent cortex, tumor, or background gray matter, thus accommodating varying tumor characteristics across patients. Our loss function also reflects the large class-imbalance present in our data. We evaluate our GNN on rs-fMRI data from 60 brain tumor patients with different tumor sizes and locations. We use motor and language task fMRI for validation. Our model achieves better localization than linear SVM, random forest, and a multilayer perceptron architecture. Our GNN is able to correctly identify bilateral language areas in the brain even when trained on patients whose language network is lateralized to the left hemisphere.
KW - Graph Neural Network
KW - Language localization
KW - Rs-fMRI
UR - http://www.scopus.com/inward/record.url?scp=85075681632&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85075681632&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-32391-2_2
DO - 10.1007/978-3-030-32391-2_2
M3 - Conference contribution
AN - SCOPUS:85075681632
SN - 9783030323905
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 10
EP - 20
BT - Connectomics in NeuroImaging - 3rd International Workshop, CNI 2019, Held in Conjunction with MICCAI 2019, Proceedings
A2 - Schirmer, Markus D.
A2 - Chung, Ai Wern
A2 - Venkataraman, Archana
A2 - Rekik, Islem
A2 - Kim, Minjeong
PB - Springer
T2 - 3rd International Workshop on Connectomics in NeuroImaging, CNI 2019 held in conjunction with the 22nd International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2019
Y2 - 13 October 2019 through 13 October 2019
ER -