Multi-task diagnosis for autism spectrum disorders using multi-modality features: A multi-center study

Jun Wang, Qian Wang, Jialin Peng, Dong Nie, Feng Zhao, Minjeong Kim, Han Zhang, Chong Yaw Wee, Shitong Wang, Dinggang Shen

Research output: Contribution to journalArticle

25 Citations (Scopus)

Abstract

Autism spectrum disorder (ASD) is a neurodevelopment disease characterized by impairment of social interaction, language, behavior, and cognitive functions. Up to now, many imaging-based methods for ASD diagnosis have been developed. For example, one may extract abundant features from multi-modality images and then derive a discriminant function to map the selected features toward the disease label. A lot of recent works, however, are limited to single imaging centers. To this end, we propose a novel multi-modality multi-center classification (M3CC) method for ASD diagnosis. We treat the classification of each imaging center as one task. By introducing the task-task and modality-modality regularizations, we solve the classification for all imaging centers simultaneously. Meanwhile, the optimal feature selection and the modeling of the discriminant functions can be jointly conducted for highly accurate diagnosis. Besides, we also present an efficient iterative optimization solution to our formulated problem and further investigate its convergence. Our comprehensive experiments on the ABIDE database show that our proposed method can significantly improve the performance of ASD diagnosis, compared to the existing methods.

Original languageEnglish
JournalHuman Brain Mapping
DOIs
Publication statusAccepted/In press - 2017

Fingerprint

Interpersonal Relations
Cognition
Language
Databases
Autism Spectrum Disorder

Keywords

  • Autism spectrum disorders
  • Feature selection
  • Modality-modality relation
  • Multi-modality data
  • Multitask learning
  • Task-task relation

ASJC Scopus subject areas

  • Anatomy
  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging
  • Neurology
  • Clinical Neurology

Cite this

Multi-task diagnosis for autism spectrum disorders using multi-modality features : A multi-center study. / Wang, Jun; Wang, Qian; Peng, Jialin; Nie, Dong; Zhao, Feng; Kim, Minjeong; Zhang, Han; Wee, Chong Yaw; Wang, Shitong; Shen, Dinggang.

In: Human Brain Mapping, 2017.

Research output: Contribution to journalArticle

Wang, Jun ; Wang, Qian ; Peng, Jialin ; Nie, Dong ; Zhao, Feng ; Kim, Minjeong ; Zhang, Han ; Wee, Chong Yaw ; Wang, Shitong ; Shen, Dinggang. / Multi-task diagnosis for autism spectrum disorders using multi-modality features : A multi-center study. In: Human Brain Mapping. 2017.
@article{e6389fb0de1c40d99a7d6443fed9967c,
title = "Multi-task diagnosis for autism spectrum disorders using multi-modality features: A multi-center study",
abstract = "Autism spectrum disorder (ASD) is a neurodevelopment disease characterized by impairment of social interaction, language, behavior, and cognitive functions. Up to now, many imaging-based methods for ASD diagnosis have been developed. For example, one may extract abundant features from multi-modality images and then derive a discriminant function to map the selected features toward the disease label. A lot of recent works, however, are limited to single imaging centers. To this end, we propose a novel multi-modality multi-center classification (M3CC) method for ASD diagnosis. We treat the classification of each imaging center as one task. By introducing the task-task and modality-modality regularizations, we solve the classification for all imaging centers simultaneously. Meanwhile, the optimal feature selection and the modeling of the discriminant functions can be jointly conducted for highly accurate diagnosis. Besides, we also present an efficient iterative optimization solution to our formulated problem and further investigate its convergence. Our comprehensive experiments on the ABIDE database show that our proposed method can significantly improve the performance of ASD diagnosis, compared to the existing methods.",
keywords = "Autism spectrum disorders, Feature selection, Modality-modality relation, Multi-modality data, Multitask learning, Task-task relation",
author = "Jun Wang and Qian Wang and Jialin Peng and Dong Nie and Feng Zhao and Minjeong Kim and Han Zhang and Wee, {Chong Yaw} and Shitong Wang and Dinggang Shen",
year = "2017",
doi = "10.1002/hbm.23575",
language = "English",
journal = "Human Brain Mapping",
issn = "1065-9471",
publisher = "Wiley-Liss Inc.",

}

TY - JOUR

T1 - Multi-task diagnosis for autism spectrum disorders using multi-modality features

T2 - A multi-center study

AU - Wang, Jun

AU - Wang, Qian

AU - Peng, Jialin

AU - Nie, Dong

AU - Zhao, Feng

AU - Kim, Minjeong

AU - Zhang, Han

AU - Wee, Chong Yaw

AU - Wang, Shitong

AU - Shen, Dinggang

PY - 2017

Y1 - 2017

N2 - Autism spectrum disorder (ASD) is a neurodevelopment disease characterized by impairment of social interaction, language, behavior, and cognitive functions. Up to now, many imaging-based methods for ASD diagnosis have been developed. For example, one may extract abundant features from multi-modality images and then derive a discriminant function to map the selected features toward the disease label. A lot of recent works, however, are limited to single imaging centers. To this end, we propose a novel multi-modality multi-center classification (M3CC) method for ASD diagnosis. We treat the classification of each imaging center as one task. By introducing the task-task and modality-modality regularizations, we solve the classification for all imaging centers simultaneously. Meanwhile, the optimal feature selection and the modeling of the discriminant functions can be jointly conducted for highly accurate diagnosis. Besides, we also present an efficient iterative optimization solution to our formulated problem and further investigate its convergence. Our comprehensive experiments on the ABIDE database show that our proposed method can significantly improve the performance of ASD diagnosis, compared to the existing methods.

AB - Autism spectrum disorder (ASD) is a neurodevelopment disease characterized by impairment of social interaction, language, behavior, and cognitive functions. Up to now, many imaging-based methods for ASD diagnosis have been developed. For example, one may extract abundant features from multi-modality images and then derive a discriminant function to map the selected features toward the disease label. A lot of recent works, however, are limited to single imaging centers. To this end, we propose a novel multi-modality multi-center classification (M3CC) method for ASD diagnosis. We treat the classification of each imaging center as one task. By introducing the task-task and modality-modality regularizations, we solve the classification for all imaging centers simultaneously. Meanwhile, the optimal feature selection and the modeling of the discriminant functions can be jointly conducted for highly accurate diagnosis. Besides, we also present an efficient iterative optimization solution to our formulated problem and further investigate its convergence. Our comprehensive experiments on the ABIDE database show that our proposed method can significantly improve the performance of ASD diagnosis, compared to the existing methods.

KW - Autism spectrum disorders

KW - Feature selection

KW - Modality-modality relation

KW - Multi-modality data

KW - Multitask learning

KW - Task-task relation

UR - http://www.scopus.com/inward/record.url?scp=85017157931&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85017157931&partnerID=8YFLogxK

U2 - 10.1002/hbm.23575

DO - 10.1002/hbm.23575

M3 - Article

C2 - 28345269

AN - SCOPUS:85017157931

JO - Human Brain Mapping

JF - Human Brain Mapping

SN - 1065-9471

ER -