Evaluating 35 methods to generate structural connectomes using pairwise classification

Dmitry Petrov; Alexander Ivanov; Joshua Faskowitz; Boris Gutman; Daniel Moyer; Julio Villalon; Neda Jahanshad; Paul Thompson

doi:10.1007/978-3-319-66182-7_59

Evaluating 35 methods to generate structural connectomes using pairwise classification

Dmitry Petrov
, Alexander Ivanov
, Joshua Faskowitz
, Boris Gutman
, Daniel Moyer
, Julio Villalon
, Neda Jahanshad
, Paul Thompson

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

14 Scopus citations

Abstract

There is no consensus on how to construct structural brain networks from diffusion MRI. How variations in pre-processing steps affect network reliability and its ability to distinguish subjects remains opaque. In this work, we address this issue by comparing 35 structural connectome-building pipelines. We vary diffusion reconstruction models, tractography algorithms and parcellations. Next, we classify structural connectome pairs as either belonging to the same individual or not. Connectome weights and eight topological derivative measures form our feature set. For experiments, we use three test-retest datasets from the Consortium for Reliability and Reproducibility (CoRR) comprised of a total of 105 individuals. We also compare pairwise classification results to a commonly used parametric test-retest measure, Intraclass Correlation Coefficient (ICC) (Code and results are available at https://github.com/lodurality/35_methods_MICCAI_2017).

Original language	English
Title of host publication	Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings
Editors	Maxime Descoteaux, Simon Duchesne, Alfred Franz, Pierre Jannin, D. Louis Collins, Lena Maier-Hein
Publisher	Springer Verlag
Pages	515-522
Number of pages	8
ISBN (Print)	9783319661810
DOIs	https://doi.org/10.1007/978-3-319-66182-7_59
State	Published - 2017
Externally published	Yes
Event	20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017 - Quebec City, Canada Duration: 11 Sep 2017 → 13 Sep 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10433 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017
Country/Territory	Canada
City	Quebec City
Period	11/09/17 → 13/09/17

Keywords

DWI
Machine learning
Structural connectomes

Access to Document

10.1007/978-3-319-66182-7_59

Cite this

Petrov, D., Ivanov, A., Faskowitz, J., Gutman, B., Moyer, D., Villalon, J., Jahanshad, N., & Thompson, P. (2017). Evaluating 35 methods to generate structural connectomes using pairwise classification. In M. Descoteaux, S. Duchesne, A. Franz, P. Jannin, D. L. Collins, & L. Maier-Hein (Eds.), Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings (pp. 515-522). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10433 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-66182-7_59

Petrov, Dmitry ; Ivanov, Alexander ; Faskowitz, Joshua et al. / Evaluating 35 methods to generate structural connectomes using pairwise classification. Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. editor / Maxime Descoteaux ; Simon Duchesne ; Alfred Franz ; Pierre Jannin ; D. Louis Collins ; Lena Maier-Hein. Springer Verlag, 2017. pp. 515-522 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{7b1a5c57a40246a8b6e36119b4a10d26,

title = "Evaluating 35 methods to generate structural connectomes using pairwise classification",

abstract = "There is no consensus on how to construct structural brain networks from diffusion MRI. How variations in pre-processing steps affect network reliability and its ability to distinguish subjects remains opaque. In this work, we address this issue by comparing 35 structural connectome-building pipelines. We vary diffusion reconstruction models, tractography algorithms and parcellations. Next, we classify structural connectome pairs as either belonging to the same individual or not. Connectome weights and eight topological derivative measures form our feature set. For experiments, we use three test-retest datasets from the Consortium for Reliability and Reproducibility (CoRR) comprised of a total of 105 individuals. We also compare pairwise classification results to a commonly used parametric test-retest measure, Intraclass Correlation Coefficient (ICC) (Code and results are available at https://github.com/lodurality/35\_methods\_MICCAI\_2017).",

keywords = "DWI, Machine learning, Structural connectomes",

author = "Dmitry Petrov and Alexander Ivanov and Joshua Faskowitz and Boris Gutman and Daniel Moyer and Julio Villalon and Neda Jahanshad and Paul Thompson",

note = "Publisher Copyright: {\textcopyright} 2017, Springer International Publishing AG.; 20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017 ; Conference date: 11-09-2017 Through 13-09-2017",

year = "2017",

doi = "10.1007/978-3-319-66182-7\_59",

language = "English",

isbn = "9783319661810",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "515--522",

editor = "Maxime Descoteaux and Simon Duchesne and Alfred Franz and Pierre Jannin and Collins, \{D. Louis\} and Lena Maier-Hein",

booktitle = "Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings",

address = "Germany",

}

Petrov, D, Ivanov, A, Faskowitz, J, Gutman, B, Moyer, D, Villalon, J, Jahanshad, N & Thompson, P 2017, Evaluating 35 methods to generate structural connectomes using pairwise classification. in M Descoteaux, S Duchesne, A Franz, P Jannin, DL Collins & L Maier-Hein (eds), Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10433 LNCS, Springer Verlag, pp. 515-522, 20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017, Quebec City, Canada, 11/09/17. https://doi.org/10.1007/978-3-319-66182-7_59

Evaluating 35 methods to generate structural connectomes using pairwise classification. / Petrov, Dmitry; Ivanov, Alexander; Faskowitz, Joshua et al.
Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. ed. / Maxime Descoteaux; Simon Duchesne; Alfred Franz; Pierre Jannin; D. Louis Collins; Lena Maier-Hein. Springer Verlag, 2017. p. 515-522 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10433 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Evaluating 35 methods to generate structural connectomes using pairwise classification

AU - Petrov, Dmitry

AU - Ivanov, Alexander

AU - Faskowitz, Joshua

AU - Gutman, Boris

AU - Moyer, Daniel

AU - Villalon, Julio

AU - Jahanshad, Neda

AU - Thompson, Paul

PY - 2017

Y1 - 2017

N2 - There is no consensus on how to construct structural brain networks from diffusion MRI. How variations in pre-processing steps affect network reliability and its ability to distinguish subjects remains opaque. In this work, we address this issue by comparing 35 structural connectome-building pipelines. We vary diffusion reconstruction models, tractography algorithms and parcellations. Next, we classify structural connectome pairs as either belonging to the same individual or not. Connectome weights and eight topological derivative measures form our feature set. For experiments, we use three test-retest datasets from the Consortium for Reliability and Reproducibility (CoRR) comprised of a total of 105 individuals. We also compare pairwise classification results to a commonly used parametric test-retest measure, Intraclass Correlation Coefficient (ICC) (Code and results are available at https://github.com/lodurality/35_methods_MICCAI_2017).

AB - There is no consensus on how to construct structural brain networks from diffusion MRI. How variations in pre-processing steps affect network reliability and its ability to distinguish subjects remains opaque. In this work, we address this issue by comparing 35 structural connectome-building pipelines. We vary diffusion reconstruction models, tractography algorithms and parcellations. Next, we classify structural connectome pairs as either belonging to the same individual or not. Connectome weights and eight topological derivative measures form our feature set. For experiments, we use three test-retest datasets from the Consortium for Reliability and Reproducibility (CoRR) comprised of a total of 105 individuals. We also compare pairwise classification results to a commonly used parametric test-retest measure, Intraclass Correlation Coefficient (ICC) (Code and results are available at https://github.com/lodurality/35_methods_MICCAI_2017).

KW - DWI

KW - Machine learning

KW - Structural connectomes

UR - https://www.scopus.com/pages/publications/85029396593

U2 - 10.1007/978-3-319-66182-7_59

DO - 10.1007/978-3-319-66182-7_59

M3 - Conference contribution

AN - SCOPUS:85029396593

SN - 9783319661810

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

SP - 515

EP - 522

BT - Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings

A2 - Descoteaux, Maxime

A2 - Duchesne, Simon

A2 - Franz, Alfred

A2 - Jannin, Pierre

A2 - Collins, D. Louis

A2 - Maier-Hein, Lena

PB - Springer Verlag

T2 - 20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017

Y2 - 11 September 2017 through 13 September 2017

ER -

Petrov D, Ivanov A, Faskowitz J, Gutman B, Moyer D, Villalon J et al. Evaluating 35 methods to generate structural connectomes using pairwise classification. In Descoteaux M, Duchesne S, Franz A, Jannin P, Collins DL, Maier-Hein L, editors, Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. Springer Verlag. 2017. p. 515-522. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-66182-7_59

Evaluating 35 methods to generate structural connectomes using pairwise classification

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this