Mapping of the Language Network With Deep Learning.

Patrick Luckett,Eric C Leuthardt,Andy G S Daniel,Ki Yun Park,Joshua S Shimony,Abraham Z Snyder,Beau M Ances,Donna Dierker,Benjamin A Seitzman,John J Lee,Carl D Hacker

doi:10.3389/fneur.2020.00819

Patrick Luckett, Eric C Leuthardt + Show 9 more

Open Access

https://doi.org/10.3389/fneur.2020.00819

Copy DOI

Abstract

Background: Pre-surgical functional localization of eloquent cortex with task-based functional MRI (T-fMRI) is part of the current standard of care prior to resection of brain tumors. Resting state fMRI (RS-fMRI) is an alternative method currently under investigation. Here, we compare group level language localization using T-fMRI vs. RS-fMRI analyzed with 3D deep convolutional neural networks (3DCNN).Methods: We analyzed data obtained in 35 patients with brain tumors that had both language T-fMRI and RS-MRI scans during pre-surgical evaluation. The T-fMRI data were analyzed using conventional techniques. The language associated resting state network was mapped using a 3DCNN previously trained with data acquired in >2,700 normal subjects. Group level results obtained by both methods were evaluated using receiver operator characteristic analysis of probability maps of language associated regions, taking as ground truth meta-analytic maps of language T-fMRI responses generated on the Neurosynth platform.Results: Both fMRI methods localized major components of the language system (areas of Broca and Wernicke). Word-stem completion T-fMRI strongly activated Broca's area but also several task-general areas not specific to language. RS-fMRI provided a more specific representation of the language system.Conclusion: 3DCNN was able to accurately localize the language network. Additionally, 3DCNN performance was remarkably tolerant of a limited quantity of RS-fMRI data.

Highlights

In treating brain tumors, the neurosurgeon must balance the benefit of maximal tumor resection against the risk of a functional impairment consequent to more aggressive approaches
The 3D deep convolutional neural networks (3DCNN) method provides highly specific maps with large probability gradients at the margin of the language regions, as would be expected of a method trained on thousands of exemplars including millions of internal parameters
The task-based fMRI (TfMRI) experiment focused on expressive language and emphasizes Broca’s region

Summary

Introduction

The neurosurgeon must balance the benefit of maximal tumor resection against the risk of a functional impairment consequent to more aggressive approaches. These two factors, maximal resection and functional preservation, are often cited in the surgical literature as predictors of long term survival [1,2,3]. Since electrocortical stimulation carries clinical risk [6], Deep Learning Language Network preoperative mapping to optimize the intraoperative surgical approach is an effective means of preserving function. Pre-surgical functional localization of eloquent cortex with task-based functional MRI (T-fMRI) is part of the current standard of care prior to resection of brain tumors. We compare group level language localization using T-fMRI vs. RS-fMRI analyzed with 3D deep convolutional neural networks (3DCNN)

Methods

Results

Discussion

Conclusion