Autism Spectrum Disorder’s Severity Prediction Model Using Utterance Features for Automatic Diagnosis Support

Masahito Sakishita, Chihiro Ogawa, Kenji J. Tsuchiya, Toshiki Iwabuchi, Taishiro Kishimoto, Yoshinobu Kano

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

Diagnoses of autism spectrum disorder (ASD) are difficult due to difference of interviewers and environments, etc. We show relations between utterance features and ASD severity scores, which were manually given by clinical psychologists. These scores are based on the Autism Diagnostic Observation Schedule (ADOS), which is the standard metrics for symptom evaluation for subjects who are suspected as ASD. We built our original corpus where we transcribed voice records of our ADOS evaluation experiment movies. Our corpus is the world largest as speech/dialog of ASD subjects, and there has been no such ADOS corpus available in Japanese language as far as we know. We investigated relationships between ADOS scores (severity) and our utterance features, automatically estimated their scores using support vector regression (SVR). Our average estimation errors were around error rates that human ADOS experts are required not to exceed. Because our detailed analysis for each part of the ADOS test (“puzzle toy assembly + story telling� part and the “depiction of a picture� part) shows different error rates, effectiveness of our features would depend on the contents of the records. Our entire results suggest a new automatic way to assist humans’ diagnosis, which could help supporting language rehabilitation for individuals with ASD in future.

Original languageEnglish
Title of host publicationStudies in Computational Intelligence
PublisherSpringer Verlag
Pages83-95
Number of pages13
DOIs
Publication statusPublished - 2020 Jan 1

Publication series

NameStudies in Computational Intelligence
Volume843
ISSN (Print)1860-949X
ISSN (Electronic)1860-9503

Fingerprint

Patient rehabilitation
Error analysis
Experiments

Keywords

  • Autism diagnostic observation schedule (ADOS)
  • Autism spectrum disorder (ASD)
  • Corpus
  • Correlation coefficient
  • Diagnosis
  • Severity
  • Support vector regression (SVR)
  • Utterance

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Sakishita, M., Ogawa, C., Tsuchiya, K. J., Iwabuchi, T., Kishimoto, T., & Kano, Y. (2020). Autism Spectrum Disorder’s Severity Prediction Model Using Utterance Features for Automatic Diagnosis Support. In Studies in Computational Intelligence (pp. 83-95). (Studies in Computational Intelligence; Vol. 843). Springer Verlag. https://doi.org/10.1007/978-3-030-24409-5_8

Autism Spectrum Disorder’s Severity Prediction Model Using Utterance Features for Automatic Diagnosis Support. / Sakishita, Masahito; Ogawa, Chihiro; Tsuchiya, Kenji J.; Iwabuchi, Toshiki; Kishimoto, Taishiro; Kano, Yoshinobu.

Studies in Computational Intelligence. Springer Verlag, 2020. p. 83-95 (Studies in Computational Intelligence; Vol. 843).

Research output: Chapter in Book/Report/Conference proceedingChapter

Sakishita, M, Ogawa, C, Tsuchiya, KJ, Iwabuchi, T, Kishimoto, T & Kano, Y 2020, Autism Spectrum Disorder’s Severity Prediction Model Using Utterance Features for Automatic Diagnosis Support. in Studies in Computational Intelligence. Studies in Computational Intelligence, vol. 843, Springer Verlag, pp. 83-95. https://doi.org/10.1007/978-3-030-24409-5_8
Sakishita M, Ogawa C, Tsuchiya KJ, Iwabuchi T, Kishimoto T, Kano Y. Autism Spectrum Disorder’s Severity Prediction Model Using Utterance Features for Automatic Diagnosis Support. In Studies in Computational Intelligence. Springer Verlag. 2020. p. 83-95. (Studies in Computational Intelligence). https://doi.org/10.1007/978-3-030-24409-5_8
Sakishita, Masahito ; Ogawa, Chihiro ; Tsuchiya, Kenji J. ; Iwabuchi, Toshiki ; Kishimoto, Taishiro ; Kano, Yoshinobu. / Autism Spectrum Disorder’s Severity Prediction Model Using Utterance Features for Automatic Diagnosis Support. Studies in Computational Intelligence. Springer Verlag, 2020. pp. 83-95 (Studies in Computational Intelligence).
@inbook{13b2734d45c1407eb73ef65019f418ab,
title = "Autism Spectrum Disorder{\^a}€™s Severity Prediction Model Using Utterance Features for Automatic Diagnosis Support",
abstract = "Diagnoses of autism spectrum disorder (ASD) are difficult due to difference of interviewers and environments, etc. We show relations between utterance features and ASD severity scores, which were manually given by clinical psychologists. These scores are based on the Autism Diagnostic Observation Schedule (ADOS), which is the standard metrics for symptom evaluation for subjects who are suspected as ASD. We built our original corpus where we transcribed voice records of our ADOS evaluation experiment movies. Our corpus is the world largest as speech/dialog of ASD subjects, and there has been no such ADOS corpus available in Japanese language as far as we know. We investigated relationships between ADOS scores (severity) and our utterance features, automatically estimated their scores using support vector regression (SVR). Our average estimation errors were around error rates that human ADOS experts are required not to exceed. Because our detailed analysis for each part of the ADOS test ({\^a}€œpuzzle toy assembly + story telling{\^a}€� part and the {\^a}€œdepiction of a picture{\^a}€� part) shows different error rates, effectiveness of our features would depend on the contents of the records. Our entire results suggest a new automatic way to assist humans{\^a}€™ diagnosis, which could help supporting language rehabilitation for individuals with ASD in future.",
keywords = "Autism diagnostic observation schedule (ADOS), Autism spectrum disorder (ASD), Corpus, Correlation coefficient, Diagnosis, Severity, Support vector regression (SVR), Utterance",
author = "Masahito Sakishita and Chihiro Ogawa and Tsuchiya, {Kenji J.} and Toshiki Iwabuchi and Taishiro Kishimoto and Yoshinobu Kano",
year = "2020",
month = "1",
day = "1",
doi = "10.1007/978-3-030-24409-5_8",
language = "English",
series = "Studies in Computational Intelligence",
publisher = "Springer Verlag",
pages = "83--95",
booktitle = "Studies in Computational Intelligence",
address = "Germany",

}

TY - CHAP

T1 - Autism Spectrum Disorder’s Severity Prediction Model Using Utterance Features for Automatic Diagnosis Support

AU - Sakishita, Masahito

AU - Ogawa, Chihiro

AU - Tsuchiya, Kenji J.

AU - Iwabuchi, Toshiki

AU - Kishimoto, Taishiro

AU - Kano, Yoshinobu

PY - 2020/1/1

Y1 - 2020/1/1

N2 - Diagnoses of autism spectrum disorder (ASD) are difficult due to difference of interviewers and environments, etc. We show relations between utterance features and ASD severity scores, which were manually given by clinical psychologists. These scores are based on the Autism Diagnostic Observation Schedule (ADOS), which is the standard metrics for symptom evaluation for subjects who are suspected as ASD. We built our original corpus where we transcribed voice records of our ADOS evaluation experiment movies. Our corpus is the world largest as speech/dialog of ASD subjects, and there has been no such ADOS corpus available in Japanese language as far as we know. We investigated relationships between ADOS scores (severity) and our utterance features, automatically estimated their scores using support vector regression (SVR). Our average estimation errors were around error rates that human ADOS experts are required not to exceed. Because our detailed analysis for each part of the ADOS test (“puzzle toy assembly + story telling� part and the “depiction of a picture� part) shows different error rates, effectiveness of our features would depend on the contents of the records. Our entire results suggest a new automatic way to assist humans’ diagnosis, which could help supporting language rehabilitation for individuals with ASD in future.

AB - Diagnoses of autism spectrum disorder (ASD) are difficult due to difference of interviewers and environments, etc. We show relations between utterance features and ASD severity scores, which were manually given by clinical psychologists. These scores are based on the Autism Diagnostic Observation Schedule (ADOS), which is the standard metrics for symptom evaluation for subjects who are suspected as ASD. We built our original corpus where we transcribed voice records of our ADOS evaluation experiment movies. Our corpus is the world largest as speech/dialog of ASD subjects, and there has been no such ADOS corpus available in Japanese language as far as we know. We investigated relationships between ADOS scores (severity) and our utterance features, automatically estimated their scores using support vector regression (SVR). Our average estimation errors were around error rates that human ADOS experts are required not to exceed. Because our detailed analysis for each part of the ADOS test (“puzzle toy assembly + story telling� part and the “depiction of a picture� part) shows different error rates, effectiveness of our features would depend on the contents of the records. Our entire results suggest a new automatic way to assist humans’ diagnosis, which could help supporting language rehabilitation for individuals with ASD in future.

KW - Autism diagnostic observation schedule (ADOS)

KW - Autism spectrum disorder (ASD)

KW - Corpus

KW - Correlation coefficient

KW - Diagnosis

KW - Severity

KW - Support vector regression (SVR)

KW - Utterance

UR - http://www.scopus.com/inward/record.url?scp=85070547628&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85070547628&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-24409-5_8

DO - 10.1007/978-3-030-24409-5_8

M3 - Chapter

AN - SCOPUS:85070547628

T3 - Studies in Computational Intelligence

SP - 83

EP - 95

BT - Studies in Computational Intelligence

PB - Springer Verlag

ER -