Deep Learning for Quality Control of Subcortical Brain 3D Shape Models
Abstract
We present several deep learning models for assessing the morphometric fidelity of deep grey matter region models extracted from brain MRI. We test three different convolutional neural net architectures (VGGNet, ResNet and Inception) over 2D maps of geometric features. Further, we present a novel geometry feature augmentation technique based on a parametric spherical mapping. Finally, we present an approach for model decision visualization, allowing human raters to see the areas of subcortical shapes most likely to be deemed of failing quality by the machine. Our training data is comprised of 5200 subjects from the ENIGMA Schizophrenia MRI cohorts, and our test dataset contains 1500 subjects from the ENIGMA Major Depressive Disorder cohorts. Our final models reduce human rater time by 46-70%. ResNet outperforms VGGNet and Inception for all of our predictive tasks.
Cite
@article{arxiv.1808.10315,
title = {Deep Learning for Quality Control of Subcortical Brain 3D Shape Models},
author = {Dmitry Petrov and Boris A. Gutman Egor Kuznetsov and Theo G. M. van Erp and Jessica A. Turner and Lianne Schmaal and Dick Veltman and Lei Wang and Kathryn Alpert and Dmitry Isaev and Artemis Zavaliangos-Petropulu and Christopher R. K. Ching and Vince Calhoun and David Glahn and Theodore D. Satterthwaite and Ole Andreas Andreassen and Stefan Borgwardt and Fleur Howells and Nynke Groenewold and Aristotle Voineskos and Joaquim Radua and Steven G. Potkin and Benedicto Crespo-Facorro and Diana Tordesillas-Gutierrez and Li Shen and Irina Lebedeva and Gianfranco Spalletta and Gary Donohoe and Peter Kochunov and Pedro G. P. Rosa and Anthony James and Udo Dannlowski and Bernhard T. Baune and Andre Aleman and Ian H. Gotlib and Henrik Walter and Martin Walter and Jair C. Soares and Stefan Ehrlich and Ruben C. Gur and N. Trung Doan and Ingrid Agartz and Lars T. Westlye and Fabienne Harrisberger and Anita Riecher-Rossler and Anne Uhlmann and Dan J. Stein and Erin W. Dickie and Edith Pomarol-Clotet and Paola Fuentes-Claramonte and Erick Jorge Canales-Rodriguez and Raymond Salvador and Alexander J. Huang and Roberto Roiz-Santianez and Shan Cong and Alexander Tomyshev and Fabrizio Piras and Daniela Vecchio and Nerisa Banaj and Valentina Ciullo and Elliot Hong and Geraldo Busatto and Marcus V. Zanetti and Mauricio H. Serpa and Simon Cervenka and Sinead Kelly and Dominik Grotegerd and Matthew D. Sacchet and Ilya M. Veer and Meng Li and Mon-Ju Wu and Benson Irungu and Esther Walton and Paul M. Thompson},
journal= {arXiv preprint arXiv:1808.10315},
year = {2018}
}
Comments
Accepted to Shape in Medical Imaging (ShapeMI) workshop at MICCAI 2018. arXiv admin note: substantial text overlap with arXiv:1707.06353