Deep Learning for Quality Control of Subcortical Brain 3D Shape Models

Dmitry Petrov, Boris A. Gutman, Egor Kuznetsov, Christopher R. K. Ching, Kathryn Alpert, Artemis Zavaliangos-Petropulu, Dmitry Isaev, Jessica A. Turner, Theo G. M. van Erp, Lei Wang, Lianne Schmaal, Dick Veltman, Paul M. Thompson

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

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 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.
Original languageEnglish
Title of host publicationShape in Medical Imaging - International Workshop, ShapeMI 2018, Held in Conjunction with MICCAI 2018, Proceedings
EditorsHervé Lombaert, Beatriz Paniagua, Bernhard Egger, Marcel Lüthi, Martin Reuter, Christian Wachinger
PublisherSpringer Verlag
Pages268-276
Volume11167 LNCS
ISBN (Print)9783030047467
DOIs
Publication statusPublished - 2018
EventInternational Workshop on Shape in Medical Imaging, ShapeMI 2018 held in conjunction with 21st International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2018 - Granada, Spain
Duration: 20 Sep 201820 Sep 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceInternational Workshop on Shape in Medical Imaging, ShapeMI 2018 held in conjunction with 21st International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2018
CountrySpain
CityGranada
Period20/09/201820/09/2018

Cite this

Petrov, D., Gutman, B. A., Kuznetsov, E., Ching, C. R. K., Alpert, K., Zavaliangos-Petropulu, A., ... Thompson, P. M. (2018). Deep Learning for Quality Control of Subcortical Brain 3D Shape Models. In H. Lombaert, B. Paniagua, B. Egger, M. Lüthi, M. Reuter, & C. Wachinger (Eds.), Shape in Medical Imaging - International Workshop, ShapeMI 2018, Held in Conjunction with MICCAI 2018, Proceedings (Vol. 11167 LNCS, pp. 268-276). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). Springer Verlag. https://doi.org/10.1007/978-3-030-04747-4_25
Petrov, Dmitry ; Gutman, Boris A. ; Kuznetsov, Egor ; Ching, Christopher R. K. ; Alpert, Kathryn ; Zavaliangos-Petropulu, Artemis ; Isaev, Dmitry ; Turner, Jessica A. ; van Erp, Theo G. M. ; Wang, Lei ; Schmaal, Lianne ; Veltman, Dick ; Thompson, Paul M. / Deep Learning for Quality Control of Subcortical Brain 3D Shape Models. Shape in Medical Imaging - International Workshop, ShapeMI 2018, Held in Conjunction with MICCAI 2018, Proceedings. editor / Hervé Lombaert ; Beatriz Paniagua ; Bernhard Egger ; Marcel Lüthi ; Martin Reuter ; Christian Wachinger. Vol. 11167 LNCS Springer Verlag, 2018. pp. 268-276 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{9f98c158e8d64218b284c8649ed4ad4d,
title = "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 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.",
author = "Dmitry Petrov and Gutman, {Boris A.} and Egor Kuznetsov and Ching, {Christopher R. K.} and Kathryn Alpert and Artemis Zavaliangos-Petropulu and Dmitry Isaev and Turner, {Jessica A.} and {van Erp}, {Theo G. M.} and Lei Wang and Lianne Schmaal and Dick Veltman and Thompson, {Paul M.}",
year = "2018",
doi = "10.1007/978-3-030-04747-4_25",
language = "English",
isbn = "9783030047467",
volume = "11167 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "268--276",
editor = "Herv{\'e} Lombaert and Beatriz Paniagua and Bernhard Egger and Marcel L{\"u}thi and Martin Reuter and Christian Wachinger",
booktitle = "Shape in Medical Imaging - International Workshop, ShapeMI 2018, Held in Conjunction with MICCAI 2018, Proceedings",

}

Petrov, D, Gutman, BA, Kuznetsov, E, Ching, CRK, Alpert, K, Zavaliangos-Petropulu, A, Isaev, D, Turner, JA, van Erp, TGM, Wang, L, Schmaal, L, Veltman, D & Thompson, PM 2018, Deep Learning for Quality Control of Subcortical Brain 3D Shape Models. in H Lombaert, B Paniagua, B Egger, M Lüthi, M Reuter & C Wachinger (eds), Shape in Medical Imaging - International Workshop, ShapeMI 2018, Held in Conjunction with MICCAI 2018, Proceedings. vol. 11167 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Springer Verlag, pp. 268-276, International Workshop on Shape in Medical Imaging, ShapeMI 2018 held in conjunction with 21st International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2018, Granada, Spain, 20/09/2018. https://doi.org/10.1007/978-3-030-04747-4_25

Deep Learning for Quality Control of Subcortical Brain 3D Shape Models. / Petrov, Dmitry; Gutman, Boris A.; Kuznetsov, Egor; Ching, Christopher R. K.; Alpert, Kathryn; Zavaliangos-Petropulu, Artemis; Isaev, Dmitry; Turner, Jessica A.; van Erp, Theo G. M.; Wang, Lei; Schmaal, Lianne; Veltman, Dick; Thompson, Paul M.

Shape in Medical Imaging - International Workshop, ShapeMI 2018, Held in Conjunction with MICCAI 2018, Proceedings. ed. / Hervé Lombaert; Beatriz Paniagua; Bernhard Egger; Marcel Lüthi; Martin Reuter; Christian Wachinger. Vol. 11167 LNCS Springer Verlag, 2018. p. 268-276 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Deep Learning for Quality Control of Subcortical Brain 3D Shape Models

AU - Petrov, Dmitry

AU - Gutman, Boris A.

AU - Kuznetsov, Egor

AU - Ching, Christopher R. K.

AU - Alpert, Kathryn

AU - Zavaliangos-Petropulu, Artemis

AU - Isaev, Dmitry

AU - Turner, Jessica A.

AU - van Erp, Theo G. M.

AU - Wang, Lei

AU - Schmaal, Lianne

AU - Veltman, Dick

AU - Thompson, Paul M.

PY - 2018

Y1 - 2018

N2 - 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 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.

AB - 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 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.

UR - https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85057390144&origin=inward

U2 - 10.1007/978-3-030-04747-4_25

DO - 10.1007/978-3-030-04747-4_25

M3 - Conference contribution

SN - 9783030047467

VL - 11167 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 268

EP - 276

BT - Shape in Medical Imaging - International Workshop, ShapeMI 2018, Held in Conjunction with MICCAI 2018, Proceedings

A2 - Lombaert, Hervé

A2 - Paniagua, Beatriz

A2 - Egger, Bernhard

A2 - Lüthi, Marcel

A2 - Reuter, Martin

A2 - Wachinger, Christian

PB - Springer Verlag

ER -

Petrov D, Gutman BA, Kuznetsov E, Ching CRK, Alpert K, Zavaliangos-Petropulu A et al. Deep Learning for Quality Control of Subcortical Brain 3D Shape Models. In Lombaert H, Paniagua B, Egger B, Lüthi M, Reuter M, Wachinger C, editors, Shape in Medical Imaging - International Workshop, ShapeMI 2018, Held in Conjunction with MICCAI 2018, Proceedings. Vol. 11167 LNCS. Springer Verlag. 2018. p. 268-276. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-04747-4_25