Feature fusion via hierarchical supervised local CCA for diagnosis of autism spectrum disorder

Feng Zhao, Lishan Qiao, Feng Shi, Pew Thian Yap, Dinggang Shen

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Early diagnosis of autism spectrum disorder (ASD) is critical for timely medical intervention, for improving patient quality of life, and for reducing the financial burden borne by the society. A key issue in neuroimaging-based ASD diagnosis is the identification of discriminating features and then fusing them to produce accurate diagnosis. In this paper, we propose a novel framework for fusing complementary and discriminating features from different imaging modalities. Specifically, we integrate the Fisher discriminant criterion and local correlation information into the canonical correlation analysis (CCA) framework, giving a new feature fusion method, called Supervised Local CCA (SL-CCA), which caters specifically to local and global multimodal features. To alleviate the neighborhood selection problem associated with SL-CCA, we further propose a hierarchical SL-CCA (HSL-CCA), by performing SL-CCA with the gradually varying neighborhood sizes. Extensive experiments on the multimodal ABIDE database show that the proposed method achieves superior performance. In addition, based on feature weight analysis, we found that only a few specific brain regions play active roles in ASD diagnosis. These brain regions include the putamen, precuneus, and orbitofrontal cortex, which are highly associated with human emotional modulation and memory formation. These finding are consistent with the behavioral phenotype of ASD.

Original languageEnglish
Pages (from-to)1-11
Number of pages11
JournalBrain Imaging and Behavior
DOIs
Publication statusAccepted/In press - 2016 Aug 17

Fingerprint

Parietal Lobe
Putamen
Brain
Prefrontal Cortex
Neuroimaging
Early Diagnosis
Quality of Life
Databases
Phenotype
Weights and Measures
Autism Spectrum Disorder

Keywords

  • Autism spectrum disorder (ASD)
  • Canonical correlation analysis (CCA)
  • Feature fusion

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Behavioral Neuroscience
  • Cellular and Molecular Neuroscience
  • Cognitive Neuroscience
  • Neurology
  • Psychiatry and Mental health
  • Clinical Neurology

Cite this

Feature fusion via hierarchical supervised local CCA for diagnosis of autism spectrum disorder. / Zhao, Feng; Qiao, Lishan; Shi, Feng; Yap, Pew Thian; Shen, Dinggang.

In: Brain Imaging and Behavior, 17.08.2016, p. 1-11.

Research output: Contribution to journalArticle

@article{f85aa08b561d4d7ebdb9279f462f32c8,
title = "Feature fusion via hierarchical supervised local CCA for diagnosis of autism spectrum disorder",
abstract = "Early diagnosis of autism spectrum disorder (ASD) is critical for timely medical intervention, for improving patient quality of life, and for reducing the financial burden borne by the society. A key issue in neuroimaging-based ASD diagnosis is the identification of discriminating features and then fusing them to produce accurate diagnosis. In this paper, we propose a novel framework for fusing complementary and discriminating features from different imaging modalities. Specifically, we integrate the Fisher discriminant criterion and local correlation information into the canonical correlation analysis (CCA) framework, giving a new feature fusion method, called Supervised Local CCA (SL-CCA), which caters specifically to local and global multimodal features. To alleviate the neighborhood selection problem associated with SL-CCA, we further propose a hierarchical SL-CCA (HSL-CCA), by performing SL-CCA with the gradually varying neighborhood sizes. Extensive experiments on the multimodal ABIDE database show that the proposed method achieves superior performance. In addition, based on feature weight analysis, we found that only a few specific brain regions play active roles in ASD diagnosis. These brain regions include the putamen, precuneus, and orbitofrontal cortex, which are highly associated with human emotional modulation and memory formation. These finding are consistent with the behavioral phenotype of ASD.",
keywords = "Autism spectrum disorder (ASD), Canonical correlation analysis (CCA), Feature fusion",
author = "Feng Zhao and Lishan Qiao and Feng Shi and Yap, {Pew Thian} and Dinggang Shen",
year = "2016",
month = "8",
day = "17",
doi = "10.1007/s11682-016-9587-5",
language = "English",
pages = "1--11",
journal = "Brain Imaging and Behavior",
issn = "1931-7557",
publisher = "Springer New York",

}

TY - JOUR

T1 - Feature fusion via hierarchical supervised local CCA for diagnosis of autism spectrum disorder

AU - Zhao, Feng

AU - Qiao, Lishan

AU - Shi, Feng

AU - Yap, Pew Thian

AU - Shen, Dinggang

PY - 2016/8/17

Y1 - 2016/8/17

N2 - Early diagnosis of autism spectrum disorder (ASD) is critical for timely medical intervention, for improving patient quality of life, and for reducing the financial burden borne by the society. A key issue in neuroimaging-based ASD diagnosis is the identification of discriminating features and then fusing them to produce accurate diagnosis. In this paper, we propose a novel framework for fusing complementary and discriminating features from different imaging modalities. Specifically, we integrate the Fisher discriminant criterion and local correlation information into the canonical correlation analysis (CCA) framework, giving a new feature fusion method, called Supervised Local CCA (SL-CCA), which caters specifically to local and global multimodal features. To alleviate the neighborhood selection problem associated with SL-CCA, we further propose a hierarchical SL-CCA (HSL-CCA), by performing SL-CCA with the gradually varying neighborhood sizes. Extensive experiments on the multimodal ABIDE database show that the proposed method achieves superior performance. In addition, based on feature weight analysis, we found that only a few specific brain regions play active roles in ASD diagnosis. These brain regions include the putamen, precuneus, and orbitofrontal cortex, which are highly associated with human emotional modulation and memory formation. These finding are consistent with the behavioral phenotype of ASD.

AB - Early diagnosis of autism spectrum disorder (ASD) is critical for timely medical intervention, for improving patient quality of life, and for reducing the financial burden borne by the society. A key issue in neuroimaging-based ASD diagnosis is the identification of discriminating features and then fusing them to produce accurate diagnosis. In this paper, we propose a novel framework for fusing complementary and discriminating features from different imaging modalities. Specifically, we integrate the Fisher discriminant criterion and local correlation information into the canonical correlation analysis (CCA) framework, giving a new feature fusion method, called Supervised Local CCA (SL-CCA), which caters specifically to local and global multimodal features. To alleviate the neighborhood selection problem associated with SL-CCA, we further propose a hierarchical SL-CCA (HSL-CCA), by performing SL-CCA with the gradually varying neighborhood sizes. Extensive experiments on the multimodal ABIDE database show that the proposed method achieves superior performance. In addition, based on feature weight analysis, we found that only a few specific brain regions play active roles in ASD diagnosis. These brain regions include the putamen, precuneus, and orbitofrontal cortex, which are highly associated with human emotional modulation and memory formation. These finding are consistent with the behavioral phenotype of ASD.

KW - Autism spectrum disorder (ASD)

KW - Canonical correlation analysis (CCA)

KW - Feature fusion

UR - http://www.scopus.com/inward/record.url?scp=84982261463&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84982261463&partnerID=8YFLogxK

U2 - 10.1007/s11682-016-9587-5

DO - 10.1007/s11682-016-9587-5

M3 - Article

C2 - 27535871

AN - SCOPUS:84982261463

SP - 1

EP - 11

JO - Brain Imaging and Behavior

JF - Brain Imaging and Behavior

SN - 1931-7557

ER -