Migraine Subclassification via a Data-Driven Automated Approach Using Multimodality Factor Mixture Modeling of Brain Structure Measurements

Todd J Schwedt, Bing Si, Jing Li, Teresa Wu, Catherine D. Chong

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Background: The current subclassification of migraine is according to headache frequency and aura status. The variability in migraine symptoms, disease course, and response to treatment suggest the presence of additional heterogeneity or subclasses within migraine. Objective: The study objective was to subclassify migraine via a data-driven approach, identifying latent factors by jointly exploiting multiple sets of brain structural features obtained via magnetic resonance imaging (MRI). Methods: Migraineurs (n = 66) and healthy controls (n = 54) had brain MRI measurements of cortical thickness, cortical surface area, and volumes for 68 regions. A multimodality factor mixture model was used to subclassify MRIs and to determine the brain structural factors that most contributed to the subclassification. Clinical characteristics of subjects in each subgroup were compared. Results: Automated MRI classification divided the subjects into two subgroups. Migraineurs in subgroup #1 had more severe allodynia symptoms during migraines (6.1 ± 5.3 vs. 3.6 ± 3.2, P =.03), more years with migraine (19.2 ± 11.3 years vs 13 ± 8.3 years, P =.01), and higher Migraine Disability Assessment (MIDAS) scores (25 ± 22.9 vs 15.7 ± 12.2, P =.04). There were not differences in headache frequency or migraine aura status between the two subgroups. Conclusions: Data-driven subclassification of brain MRIs based upon structural measurements identified two subgroups. Amongst migraineurs, the subgroups differed in allodynia symptom severity, years with migraine, and migraine-related disability. Since allodynia is associated with this imaging-based subclassification of migraine and prior publications suggest that allodynia impacts migraine treatment response and disease prognosis, future migraine diagnostic criteria could consider allodynia when defining migraine subgroups.

Original languageEnglish (US)
Pages (from-to)1051-1064
Number of pages14
JournalHeadache
Volume57
Issue number7
DOIs
StatePublished - Jul 1 2017

Fingerprint

Migraine Disorders
Brain
Hyperalgesia
Magnetic Resonance Imaging
Headache
Epilepsy
Publications

Keywords

  • allodynia
  • brain structure
  • brain volume
  • classification
  • cortical surface area
  • cortical thickness
  • factor mixture model
  • headache
  • magnetic resonance imaging
  • migraine
  • multimodality factor mixture model

ASJC Scopus subject areas

  • Neurology
  • Clinical Neurology

Cite this

Migraine Subclassification via a Data-Driven Automated Approach Using Multimodality Factor Mixture Modeling of Brain Structure Measurements. / Schwedt, Todd J; Si, Bing; Li, Jing; Wu, Teresa; Chong, Catherine D.

In: Headache, Vol. 57, No. 7, 01.07.2017, p. 1051-1064.

Research output: Contribution to journalArticle

@article{0e736899f739479d9475d1fb0088edb6,
title = "Migraine Subclassification via a Data-Driven Automated Approach Using Multimodality Factor Mixture Modeling of Brain Structure Measurements",
abstract = "Background: The current subclassification of migraine is according to headache frequency and aura status. The variability in migraine symptoms, disease course, and response to treatment suggest the presence of additional heterogeneity or subclasses within migraine. Objective: The study objective was to subclassify migraine via a data-driven approach, identifying latent factors by jointly exploiting multiple sets of brain structural features obtained via magnetic resonance imaging (MRI). Methods: Migraineurs (n = 66) and healthy controls (n = 54) had brain MRI measurements of cortical thickness, cortical surface area, and volumes for 68 regions. A multimodality factor mixture model was used to subclassify MRIs and to determine the brain structural factors that most contributed to the subclassification. Clinical characteristics of subjects in each subgroup were compared. Results: Automated MRI classification divided the subjects into two subgroups. Migraineurs in subgroup #1 had more severe allodynia symptoms during migraines (6.1 ± 5.3 vs. 3.6 ± 3.2, P =.03), more years with migraine (19.2 ± 11.3 years vs 13 ± 8.3 years, P =.01), and higher Migraine Disability Assessment (MIDAS) scores (25 ± 22.9 vs 15.7 ± 12.2, P =.04). There were not differences in headache frequency or migraine aura status between the two subgroups. Conclusions: Data-driven subclassification of brain MRIs based upon structural measurements identified two subgroups. Amongst migraineurs, the subgroups differed in allodynia symptom severity, years with migraine, and migraine-related disability. Since allodynia is associated with this imaging-based subclassification of migraine and prior publications suggest that allodynia impacts migraine treatment response and disease prognosis, future migraine diagnostic criteria could consider allodynia when defining migraine subgroups.",
keywords = "allodynia, brain structure, brain volume, classification, cortical surface area, cortical thickness, factor mixture model, headache, magnetic resonance imaging, migraine, multimodality factor mixture model",
author = "Schwedt, {Todd J} and Bing Si and Jing Li and Teresa Wu and Chong, {Catherine D.}",
year = "2017",
month = "7",
day = "1",
doi = "10.1111/head.13121",
language = "English (US)",
volume = "57",
pages = "1051--1064",
journal = "Headache",
issn = "0017-8748",
publisher = "Wiley-Blackwell",
number = "7",

}

TY - JOUR

T1 - Migraine Subclassification via a Data-Driven Automated Approach Using Multimodality Factor Mixture Modeling of Brain Structure Measurements

AU - Schwedt, Todd J

AU - Si, Bing

AU - Li, Jing

AU - Wu, Teresa

AU - Chong, Catherine D.

PY - 2017/7/1

Y1 - 2017/7/1

N2 - Background: The current subclassification of migraine is according to headache frequency and aura status. The variability in migraine symptoms, disease course, and response to treatment suggest the presence of additional heterogeneity or subclasses within migraine. Objective: The study objective was to subclassify migraine via a data-driven approach, identifying latent factors by jointly exploiting multiple sets of brain structural features obtained via magnetic resonance imaging (MRI). Methods: Migraineurs (n = 66) and healthy controls (n = 54) had brain MRI measurements of cortical thickness, cortical surface area, and volumes for 68 regions. A multimodality factor mixture model was used to subclassify MRIs and to determine the brain structural factors that most contributed to the subclassification. Clinical characteristics of subjects in each subgroup were compared. Results: Automated MRI classification divided the subjects into two subgroups. Migraineurs in subgroup #1 had more severe allodynia symptoms during migraines (6.1 ± 5.3 vs. 3.6 ± 3.2, P =.03), more years with migraine (19.2 ± 11.3 years vs 13 ± 8.3 years, P =.01), and higher Migraine Disability Assessment (MIDAS) scores (25 ± 22.9 vs 15.7 ± 12.2, P =.04). There were not differences in headache frequency or migraine aura status between the two subgroups. Conclusions: Data-driven subclassification of brain MRIs based upon structural measurements identified two subgroups. Amongst migraineurs, the subgroups differed in allodynia symptom severity, years with migraine, and migraine-related disability. Since allodynia is associated with this imaging-based subclassification of migraine and prior publications suggest that allodynia impacts migraine treatment response and disease prognosis, future migraine diagnostic criteria could consider allodynia when defining migraine subgroups.

AB - Background: The current subclassification of migraine is according to headache frequency and aura status. The variability in migraine symptoms, disease course, and response to treatment suggest the presence of additional heterogeneity or subclasses within migraine. Objective: The study objective was to subclassify migraine via a data-driven approach, identifying latent factors by jointly exploiting multiple sets of brain structural features obtained via magnetic resonance imaging (MRI). Methods: Migraineurs (n = 66) and healthy controls (n = 54) had brain MRI measurements of cortical thickness, cortical surface area, and volumes for 68 regions. A multimodality factor mixture model was used to subclassify MRIs and to determine the brain structural factors that most contributed to the subclassification. Clinical characteristics of subjects in each subgroup were compared. Results: Automated MRI classification divided the subjects into two subgroups. Migraineurs in subgroup #1 had more severe allodynia symptoms during migraines (6.1 ± 5.3 vs. 3.6 ± 3.2, P =.03), more years with migraine (19.2 ± 11.3 years vs 13 ± 8.3 years, P =.01), and higher Migraine Disability Assessment (MIDAS) scores (25 ± 22.9 vs 15.7 ± 12.2, P =.04). There were not differences in headache frequency or migraine aura status between the two subgroups. Conclusions: Data-driven subclassification of brain MRIs based upon structural measurements identified two subgroups. Amongst migraineurs, the subgroups differed in allodynia symptom severity, years with migraine, and migraine-related disability. Since allodynia is associated with this imaging-based subclassification of migraine and prior publications suggest that allodynia impacts migraine treatment response and disease prognosis, future migraine diagnostic criteria could consider allodynia when defining migraine subgroups.

KW - allodynia

KW - brain structure

KW - brain volume

KW - classification

KW - cortical surface area

KW - cortical thickness

KW - factor mixture model

KW - headache

KW - magnetic resonance imaging

KW - migraine

KW - multimodality factor mixture model

UR - http://www.scopus.com/inward/record.url?scp=85020516361&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020516361&partnerID=8YFLogxK

U2 - 10.1111/head.13121

DO - 10.1111/head.13121

M3 - Article

C2 - 28627714

AN - SCOPUS:85020516361

VL - 57

SP - 1051

EP - 1064

JO - Headache

JF - Headache

SN - 0017-8748

IS - 7

ER -