Automated cortical thickness and skewness feature selection in bipolar disorder using a semi-supervised learning method

L. Squarcina, T. M. Dagnew, M. W. Rivolta, M. Bellani, R. Sassi, P. Brambilla

Research output: Contribution to journalArticle

Abstract

Background: Bipolar disorder (BD) broadly affects brain structure, in particular areas involved in emotion processing and cognition. In the last years, the psychiatric field's interest in machine learning approaches has been steadily growing, thanks to the potentiality of automatically discriminating patients from healthy controls. Methods: In this work, we employed cortical thickness of 58 regions of interest obtained from magnetic resonance imaging scans of 41 BD patients and 34 healthy controls, to automatically identify the regions which are mostly involved with the disease. We used a semi-supervised method, addressing the criticisms on supervised methods, related to the fact that the diagnosis is not unaffected by uncertainty. Results: Our results confirm findings in previous studies, with a classification accuracy of about 75% when mean thickness and skewness of up to five regions are considered. We obtained that the parietal lobe and some areas in the temporal sulcus were the regions which were the most involved with BD. Limitations: The major limitation of our work is the limited size or our dataset, but in line with other recent machine learning works in the field. Moreover, we considered chronic patients, whose brain characteristics may thus be affected. Conclusions: The automatic selection of the brain regions most involved in BD may be of great importance when dealing with the pathogenesis of the disorder. Our method selected regions which are known to be involved with BD, indicating that damage to the identified areas can be considered as a marker of disease.

Original languageEnglish
Pages (from-to)416-423
Number of pages8
JournalJournal of Affective Disorders
Volume256
DOIs
Publication statusPublished - Sep 1 2019

Fingerprint

Bipolar Disorder
Brain
Parietal Lobe
Temporal Lobe
Cognition
Uncertainty
Psychiatry
Emotions
Magnetic Resonance Imaging
Supervised Machine Learning
Machine Learning

Keywords

  • Bipolar disorder
  • Classification
  • Machine learning
  • Magnetic resonance imaging
  • MRI pattern recognition
  • Neuroimaging
  • Semi-supervised learning

ASJC Scopus subject areas

  • Clinical Psychology
  • Psychiatry and Mental health

Cite this

Automated cortical thickness and skewness feature selection in bipolar disorder using a semi-supervised learning method. / Squarcina, L.; Dagnew, T. M.; Rivolta, M. W.; Bellani, M.; Sassi, R.; Brambilla, P.

In: Journal of Affective Disorders, Vol. 256, 01.09.2019, p. 416-423.

Research output: Contribution to journalArticle

Squarcina, L. ; Dagnew, T. M. ; Rivolta, M. W. ; Bellani, M. ; Sassi, R. ; Brambilla, P. / Automated cortical thickness and skewness feature selection in bipolar disorder using a semi-supervised learning method. In: Journal of Affective Disorders. 2019 ; Vol. 256. pp. 416-423.
@article{50e4c1e3e6fb4dbcb0ee0d7cc0c40e65,
title = "Automated cortical thickness and skewness feature selection in bipolar disorder using a semi-supervised learning method",
abstract = "Background: Bipolar disorder (BD) broadly affects brain structure, in particular areas involved in emotion processing and cognition. In the last years, the psychiatric field's interest in machine learning approaches has been steadily growing, thanks to the potentiality of automatically discriminating patients from healthy controls. Methods: In this work, we employed cortical thickness of 58 regions of interest obtained from magnetic resonance imaging scans of 41 BD patients and 34 healthy controls, to automatically identify the regions which are mostly involved with the disease. We used a semi-supervised method, addressing the criticisms on supervised methods, related to the fact that the diagnosis is not unaffected by uncertainty. Results: Our results confirm findings in previous studies, with a classification accuracy of about 75{\%} when mean thickness and skewness of up to five regions are considered. We obtained that the parietal lobe and some areas in the temporal sulcus were the regions which were the most involved with BD. Limitations: The major limitation of our work is the limited size or our dataset, but in line with other recent machine learning works in the field. Moreover, we considered chronic patients, whose brain characteristics may thus be affected. Conclusions: The automatic selection of the brain regions most involved in BD may be of great importance when dealing with the pathogenesis of the disorder. Our method selected regions which are known to be involved with BD, indicating that damage to the identified areas can be considered as a marker of disease.",
keywords = "Bipolar disorder, Classification, Machine learning, Magnetic resonance imaging, MRI pattern recognition, Neuroimaging, Semi-supervised learning",
author = "L. Squarcina and Dagnew, {T. M.} and Rivolta, {M. W.} and M. Bellani and R. Sassi and P. Brambilla",
year = "2019",
month = "9",
day = "1",
doi = "10.1016/j.jad.2019.06.019",
language = "English",
volume = "256",
pages = "416--423",
journal = "Journal of Affective Disorders",
issn = "0165-0327",
publisher = "Elsevier",

}

TY - JOUR

T1 - Automated cortical thickness and skewness feature selection in bipolar disorder using a semi-supervised learning method

AU - Squarcina, L.

AU - Dagnew, T. M.

AU - Rivolta, M. W.

AU - Bellani, M.

AU - Sassi, R.

AU - Brambilla, P.

PY - 2019/9/1

Y1 - 2019/9/1

N2 - Background: Bipolar disorder (BD) broadly affects brain structure, in particular areas involved in emotion processing and cognition. In the last years, the psychiatric field's interest in machine learning approaches has been steadily growing, thanks to the potentiality of automatically discriminating patients from healthy controls. Methods: In this work, we employed cortical thickness of 58 regions of interest obtained from magnetic resonance imaging scans of 41 BD patients and 34 healthy controls, to automatically identify the regions which are mostly involved with the disease. We used a semi-supervised method, addressing the criticisms on supervised methods, related to the fact that the diagnosis is not unaffected by uncertainty. Results: Our results confirm findings in previous studies, with a classification accuracy of about 75% when mean thickness and skewness of up to five regions are considered. We obtained that the parietal lobe and some areas in the temporal sulcus were the regions which were the most involved with BD. Limitations: The major limitation of our work is the limited size or our dataset, but in line with other recent machine learning works in the field. Moreover, we considered chronic patients, whose brain characteristics may thus be affected. Conclusions: The automatic selection of the brain regions most involved in BD may be of great importance when dealing with the pathogenesis of the disorder. Our method selected regions which are known to be involved with BD, indicating that damage to the identified areas can be considered as a marker of disease.

AB - Background: Bipolar disorder (BD) broadly affects brain structure, in particular areas involved in emotion processing and cognition. In the last years, the psychiatric field's interest in machine learning approaches has been steadily growing, thanks to the potentiality of automatically discriminating patients from healthy controls. Methods: In this work, we employed cortical thickness of 58 regions of interest obtained from magnetic resonance imaging scans of 41 BD patients and 34 healthy controls, to automatically identify the regions which are mostly involved with the disease. We used a semi-supervised method, addressing the criticisms on supervised methods, related to the fact that the diagnosis is not unaffected by uncertainty. Results: Our results confirm findings in previous studies, with a classification accuracy of about 75% when mean thickness and skewness of up to five regions are considered. We obtained that the parietal lobe and some areas in the temporal sulcus were the regions which were the most involved with BD. Limitations: The major limitation of our work is the limited size or our dataset, but in line with other recent machine learning works in the field. Moreover, we considered chronic patients, whose brain characteristics may thus be affected. Conclusions: The automatic selection of the brain regions most involved in BD may be of great importance when dealing with the pathogenesis of the disorder. Our method selected regions which are known to be involved with BD, indicating that damage to the identified areas can be considered as a marker of disease.

KW - Bipolar disorder

KW - Classification

KW - Machine learning

KW - Magnetic resonance imaging

KW - MRI pattern recognition

KW - Neuroimaging

KW - Semi-supervised learning

UR - http://www.scopus.com/inward/record.url?scp=85067568930&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85067568930&partnerID=8YFLogxK

U2 - 10.1016/j.jad.2019.06.019

DO - 10.1016/j.jad.2019.06.019

M3 - Article

AN - SCOPUS:85067568930

VL - 256

SP - 416

EP - 423

JO - Journal of Affective Disorders

JF - Journal of Affective Disorders

SN - 0165-0327

ER -