Searchlight based feature extraction

Shahar Jamshy; Omri Perez; Yehezkel Yeshurun; Talma Hendler; Nathan Intrator

doi:10.1007/978-3-642-34713-9_3

Searchlight based feature extraction

Shahar Jamshy^*, Omri Perez, Yehezkel Yeshurun, Talma Hendler, Nathan Intrator

^*Corresponding author for this work

Research output: Contribution to journal › Conference article › peer-review

2 Scopus citations

Abstract

A multi voxel pattern analysis classification framework suitable for neuroimaging data is introduced. The framework includes a novel feature extraction method that uses local modeling based on domain specific knowledge, and therefore, can produce better whole-brain global classification performance using a smaller number of features. In particular, the method includes spherical searchlights in combination with local SVM modeling. The performance of the framework is demonstrated on a challenging fMRI classification problem, and is found to be superior to the performance of state-of-the-art feature selection methods used in neuroimaging.

Original language	English
Pages (from-to)	17-25
Number of pages	9
Journal	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	7263 LNAI
DOIs	https://doi.org/10.1007/978-3-642-34713-9_3
State	Published - 2012
Event	International Workshop on Machine Learning and Interpretation in Neuroimaging, MLINI 2011, Held at Neural Information Processing, NIPS 2011 - Sierra Nevada, Spain Duration: 16 Dec 2011 → 17 Dec 2011

Keywords

MVPA
SVM
Searchlight
fMRI

Access to Document

10.1007/978-3-642-34713-9_3

Cite this

@article{f6761fa2ac934256b0a01202736b682e,

title = "Searchlight based feature extraction",

abstract = "A multi voxel pattern analysis classification framework suitable for neuroimaging data is introduced. The framework includes a novel feature extraction method that uses local modeling based on domain specific knowledge, and therefore, can produce better whole-brain global classification performance using a smaller number of features. In particular, the method includes spherical searchlights in combination with local SVM modeling. The performance of the framework is demonstrated on a challenging fMRI classification problem, and is found to be superior to the performance of state-of-the-art feature selection methods used in neuroimaging.",

keywords = "MVPA, SVM, Searchlight, fMRI",

author = "Shahar Jamshy and Omri Perez and Yehezkel Yeshurun and Talma Hendler and Nathan Intrator",

year = "2012",

doi = "10.1007/978-3-642-34713-9_3",

language = "אנגלית",

volume = "7263 LNAI",

pages = "17--25",

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

issn = "0302-9743",

publisher = "Springer Science and Business Media Deutschland GmbH",

note = "International Workshop on Machine Learning and Interpretation in Neuroimaging, MLINI 2011, Held at Neural Information Processing, NIPS 2011 ; Conference date: 16-12-2011 Through 17-12-2011",

}

TY - JOUR

T1 - Searchlight based feature extraction

AU - Jamshy, Shahar

AU - Perez, Omri

AU - Yeshurun, Yehezkel

AU - Hendler, Talma

AU - Intrator, Nathan

PY - 2012

Y1 - 2012

N2 - A multi voxel pattern analysis classification framework suitable for neuroimaging data is introduced. The framework includes a novel feature extraction method that uses local modeling based on domain specific knowledge, and therefore, can produce better whole-brain global classification performance using a smaller number of features. In particular, the method includes spherical searchlights in combination with local SVM modeling. The performance of the framework is demonstrated on a challenging fMRI classification problem, and is found to be superior to the performance of state-of-the-art feature selection methods used in neuroimaging.

AB - A multi voxel pattern analysis classification framework suitable for neuroimaging data is introduced. The framework includes a novel feature extraction method that uses local modeling based on domain specific knowledge, and therefore, can produce better whole-brain global classification performance using a smaller number of features. In particular, the method includes spherical searchlights in combination with local SVM modeling. The performance of the framework is demonstrated on a challenging fMRI classification problem, and is found to be superior to the performance of state-of-the-art feature selection methods used in neuroimaging.

KW - MVPA

KW - SVM

KW - Searchlight

KW - fMRI

UR - http://www.scopus.com/inward/record.url?scp=84869994507&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-34713-9_3

DO - 10.1007/978-3-642-34713-9_3

M3 - ???researchoutput.researchoutputtypes.contributiontojournal.conferencearticle???

AN - SCOPUS:84869994507

SN - 0302-9743

VL - 7263 LNAI

SP - 17

EP - 25

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

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

T2 - International Workshop on Machine Learning and Interpretation in Neuroimaging, MLINI 2011, Held at Neural Information Processing, NIPS 2011

Y2 - 16 December 2011 through 17 December 2011

ER -

Searchlight based feature extraction

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this