• HOME
  • News
  • Development of a schizophrenia-specific classifier using machine learning of brain image data
OUTCOMES
2022.04.12

Development of a schizophrenia-specific classifier using machine learning of brain image data

Summary

A research group led by Dr. Koike, Prof. Kasai, Prof. Abe (The University of Tokyo), and Prof. Yamasue (Hamamatsu University) has performed machine learning using multiple structural brain MRI datasets measured from chronic-phase schizophrenia and healthy controls, and developed a classifier. This classifier is capable of discriminating between schizophrenia and healthy controls when fitted with brain imaging data from different clinical stages of schizophrenia (high risk for psychosis, first episode psychosis) and developmental disorder groups. The probability of being discriminated as chronic schizophrenia was higher for first-episode psychosis than for psychosis high-risk, and more than 80% of the developmental disability group were discriminated as healthy controls. Therefore, this classifier is expected to be applied as a marker for differential diagnosis and treatment prediction necessary in clinical practice.

【Information on the outcome】

https://www.amed.go.jp/news/release_20210907.html (AMED)
https://www.u-tokyo.ac.jp/focus/ja/press/z0109_00039.html# (The University of Tokyo)

Article

<Title>

Application of a machine learning algorithm for structural brain images in chronic schizophrenia to earlier clinical stages of psychosis and autism spectrum disorder: A multi-protocol imaging dataset study
DOI : 10.1093/schbul/sbac030

<Authors>

Yinghan Zhu, Hironori Nakatani, Walid Yassin, Norihide Maikusa, Naohiro Okada, Akira Kunimatsu, Osamu Abe, Hitoshi Kuwabara, Hidenori Yamasue, Kiyoto Kasai, Kazuo Okanoya, Shinsuke Koike

<Journal>

Schizophrenia Bulletin