Complexity in mood disorder diagnosis: fMRI connectivity networks predicted medication-class of response in complex patients

E. Osuch, S. Gao, M. Wammes, J. Théberge, P. Willimason, R. J. Neufeld, Y. Du, J. Sui, Vince Daniel Calhoun

Research output: Contribution to journalArticle

Abstract

Objective: This study determined the clinical utility of an fMRI classification algorithm predicting medication-class of response in patients with challenging mood diagnoses. Methods: Ninety-nine 16–27-year-olds underwent resting state fMRI scans in three groups—BD, MDD and healthy controls. A predictive algorithm was trained and cross-validated on the known-diagnosis patients using maximally spatially independent components (ICs), constructing a similarity matrix among subjects, partitioning the matrix in kernel space and optimizing support vector machine classifiers and IC combinations. This classifier was also applied to each of 12 new individual patients with unclear mood disorder diagnoses. Results: Classification within the known-diagnosis group was approximately 92.4% accurate. The five maximally contributory ICs were identified. Applied to the complicated patients, the algorithm diagnosis was consistent with optimal medication-class of response to sustained recovery in 11 of 12 cases (i.e., almost 92% accuracy). Conclusion: This classification algorithm performed well for the know-diagnosis but also predicted medication-class of response in difficult-to-diagnose patients. Further research can enhance this approach and extend these findings to be more clinically accessible.

Original languageEnglish (US)
JournalActa Psychiatrica Scandinavica
DOIs
StateAccepted/In press - Jan 1 2018
Externally publishedYes

Fingerprint

Mood Disorders
Magnetic Resonance Imaging
Research

Keywords

  • bipolar disorder
  • differential diagnosis
  • functional neuroimaging
  • machine learning
  • mood disorders

ASJC Scopus subject areas

  • Psychiatry and Mental health

Cite this

Complexity in mood disorder diagnosis : fMRI connectivity networks predicted medication-class of response in complex patients. / Osuch, E.; Gao, S.; Wammes, M.; Théberge, J.; Willimason, P.; Neufeld, R. J.; Du, Y.; Sui, J.; Calhoun, Vince Daniel.

In: Acta Psychiatrica Scandinavica, 01.01.2018.

Research output: Contribution to journalArticle

@article{a42e3def3fea4a6b9340dc6cae9be9d8,
title = "Complexity in mood disorder diagnosis: fMRI connectivity networks predicted medication-class of response in complex patients",
abstract = "Objective: This study determined the clinical utility of an fMRI classification algorithm predicting medication-class of response in patients with challenging mood diagnoses. Methods: Ninety-nine 16–27-year-olds underwent resting state fMRI scans in three groups—BD, MDD and healthy controls. A predictive algorithm was trained and cross-validated on the known-diagnosis patients using maximally spatially independent components (ICs), constructing a similarity matrix among subjects, partitioning the matrix in kernel space and optimizing support vector machine classifiers and IC combinations. This classifier was also applied to each of 12 new individual patients with unclear mood disorder diagnoses. Results: Classification within the known-diagnosis group was approximately 92.4{\%} accurate. The five maximally contributory ICs were identified. Applied to the complicated patients, the algorithm diagnosis was consistent with optimal medication-class of response to sustained recovery in 11 of 12 cases (i.e., almost 92{\%} accuracy). Conclusion: This classification algorithm performed well for the know-diagnosis but also predicted medication-class of response in difficult-to-diagnose patients. Further research can enhance this approach and extend these findings to be more clinically accessible.",
keywords = "bipolar disorder, differential diagnosis, functional neuroimaging, machine learning, mood disorders",
author = "E. Osuch and S. Gao and M. Wammes and J. Th{\'e}berge and P. Willimason and Neufeld, {R. J.} and Y. Du and J. Sui and Calhoun, {Vince Daniel}",
year = "2018",
month = "1",
day = "1",
doi = "10.1111/acps.12945",
language = "English (US)",
journal = "Acta Psychiatrica Scandinavica",
issn = "0001-690X",
publisher = "Wiley-Blackwell",

}

TY - JOUR

T1 - Complexity in mood disorder diagnosis

T2 - fMRI connectivity networks predicted medication-class of response in complex patients

AU - Osuch, E.

AU - Gao, S.

AU - Wammes, M.

AU - Théberge, J.

AU - Willimason, P.

AU - Neufeld, R. J.

AU - Du, Y.

AU - Sui, J.

AU - Calhoun, Vince Daniel

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Objective: This study determined the clinical utility of an fMRI classification algorithm predicting medication-class of response in patients with challenging mood diagnoses. Methods: Ninety-nine 16–27-year-olds underwent resting state fMRI scans in three groups—BD, MDD and healthy controls. A predictive algorithm was trained and cross-validated on the known-diagnosis patients using maximally spatially independent components (ICs), constructing a similarity matrix among subjects, partitioning the matrix in kernel space and optimizing support vector machine classifiers and IC combinations. This classifier was also applied to each of 12 new individual patients with unclear mood disorder diagnoses. Results: Classification within the known-diagnosis group was approximately 92.4% accurate. The five maximally contributory ICs were identified. Applied to the complicated patients, the algorithm diagnosis was consistent with optimal medication-class of response to sustained recovery in 11 of 12 cases (i.e., almost 92% accuracy). Conclusion: This classification algorithm performed well for the know-diagnosis but also predicted medication-class of response in difficult-to-diagnose patients. Further research can enhance this approach and extend these findings to be more clinically accessible.

AB - Objective: This study determined the clinical utility of an fMRI classification algorithm predicting medication-class of response in patients with challenging mood diagnoses. Methods: Ninety-nine 16–27-year-olds underwent resting state fMRI scans in three groups—BD, MDD and healthy controls. A predictive algorithm was trained and cross-validated on the known-diagnosis patients using maximally spatially independent components (ICs), constructing a similarity matrix among subjects, partitioning the matrix in kernel space and optimizing support vector machine classifiers and IC combinations. This classifier was also applied to each of 12 new individual patients with unclear mood disorder diagnoses. Results: Classification within the known-diagnosis group was approximately 92.4% accurate. The five maximally contributory ICs were identified. Applied to the complicated patients, the algorithm diagnosis was consistent with optimal medication-class of response to sustained recovery in 11 of 12 cases (i.e., almost 92% accuracy). Conclusion: This classification algorithm performed well for the know-diagnosis but also predicted medication-class of response in difficult-to-diagnose patients. Further research can enhance this approach and extend these findings to be more clinically accessible.

KW - bipolar disorder

KW - differential diagnosis

KW - functional neuroimaging

KW - machine learning

KW - mood disorders

UR - http://www.scopus.com/inward/record.url?scp=85052614335&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85052614335&partnerID=8YFLogxK

U2 - 10.1111/acps.12945

DO - 10.1111/acps.12945

M3 - Article

C2 - 30084192

AN - SCOPUS:85052614335

JO - Acta Psychiatrica Scandinavica

JF - Acta Psychiatrica Scandinavica

SN - 0001-690X

ER -