Spectral Clustering Based on Structural Magnetic Resonance Imaging and its Relationship with Major Depressive Disorder and Cognitive Ability

Hon Wah Yeung; Xueyi Shen; Aleks Stolicyn; Laura De Nooij; Mathew A Harris; Liana Romaniuk; Colin R Buchanan; Gordon D Waiter; Anca-Larisa Sandu; Christopher J McNeil; Alison Murray; J Douglas Steele; Archie Campbell; David Porteous; Stephen M Lawrie; Andrew M McIntosh; Simon R Cox; Keith M Smith; Heather C Whalley

doi:10.1111/ejn.15423

Spectral Clustering Based on Structural Magnetic Resonance Imaging and its Relationship with Major Depressive Disorder and Cognitive Ability

Hon Wah Yeung^* (Corresponding Author), Xueyi Shen, Aleks Stolicyn, Laura De Nooij, Mathew A Harris, Liana Romaniuk, Colin R Buchanan, Gordon D Waiter, Anca-Larisa Sandu, Christopher J McNeil, Alison Murray, J Douglas Steele, Archie Campbell, David Porteous, Stephen M Lawrie, Andrew M McIntosh, Simon R Cox, Keith M Smith, Heather C Whalley

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

4 Citations (Scopus)

3 Downloads (Pure)

Abstract

There is increasing interest in using data-driven unsupervised methods to identify structural underpinnings of common mental illnesses, including Major Depressive Disorder (MDD) and associated traits such as cognition. However, studies are often limited to severe clinical cases with small sample sizes and most do not include replication. Here, we examine two relatively large samples with structural magnetic resonance imaging (MRI), measures of lifetime MDD and cognitive variables: Generation Scotland (GS subsample, N = 980), and UK Biobank (UKB, N = 8900); for discovery and replication, using an exploratory approach. Regional measures of FreeSurfer derived cortical thickness (CT), cortical surface area (CSA), cortical volume (CV) and subcortical volume (subCV) were input into a clustering process, controlling for common covariates. The main analysis steps involved constructing participant K-nearest neighbour graphs, and graph partitioning with Markov Stability to determine optimal clustering of participants. Resultant clusters were i) checked whether they were replicated in an independent cohort, and ii) tested for associations with depression status, and cognitive measures. Participants separated into two clusters based on structural brain measurements in GS subsample, with large Cohen's d effect sizes between clusters in higher order cortical regions, commonly associated with executive function and decision making. Clustering was replicated in the UKB sample, with high correlations of cluster effect sizes for CT, CSA, CV and subCV between cohorts across regions. The identified clusters were not significantly different with respect to MDD case-control status in either cohort (GS subsample: pFDR = 0.2239 - 0.6585; UKB: pFDR = 0.2003 - 0.7690). Significant differences in general cognitive ability were, however, found between the clusters for both datasets; for CSA, CV, subCV (GS subsample: d = 0.2529 - 0.3490, pFDR < 0.005; UKB: d = 0.0868 - 0.1070, pFDR < 0.005). Our results suggest that there are replicable natural groupings of participants based on cortical and subcortical brain measures, which may be related to differences in cognitive performance, but not to the MDD case-control status.

Original language	English
Pages (from-to)	6281-6303
Number of pages	23
Journal	European Journal of Neuroscience
Volume	54
Issue number	6
Early online date	14 Aug 2021
DOIs	https://doi.org/10.1111/ejn.15423
Publication status	Published - 30 Sept 2021

Bibliographical note

Acknowledgement
This study was supported and funded by the Wellcome Trust Strategic Award “Stratifying Resilience and Depression Longitudinally” (STRADL) (Reference 104036/Z/14/Z), and was also supported by National Institutes of Health (NIH) research grant R01AG054628. Generation Scotland received core support from the Chief Scientist Office of the Scottish Government Health Directorates [CZD/16/6] and the Scottish Funding Council [HR03006] and is currently supported by the Wellcome Trust [216767/Z/19/Z]. The research was conducted using the UKB resource, with approved project number 10279, and the UKB imaging data was processed at the University of Edinburgh Centre for Cognitive Ageing and Cognitive Epidemiology (CCACE) (http://www.ccace.ed.ac.uk), part of the cross-council Lifelong Health and Wellbeing Initiative (MR/K026992/1). CCACE received funding from Biotechnology and Biological Sciences Research Council (BBSRC), Medical Research Council (MRC), and was also supported by Age UK as part of The Disconnected Mind project. Keith M. Smith was supported by Health Data Research UK, an initiative funded by UK Research and Innovation Councils, NIH Research (England) and the UK devolved administrations, and leading medical research charities. Simon R. Cox was supported by Age UK (Disconnected Mind project), the UK Medical Research Council [MRC: MR/R024065/1], and the US National Institutes of Health (NIH) [R01AG054628].

Keywords

Major Depressive Disorder
Cognition
Machine Learning
Clustering, Markov Stability
Structural Neuroimaging

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1111/ejn.15423Licence: Unspecified

Yeung_etal_EJN_Spectral_Clustering_Based_AAM
Peer reviewed version of the following article: [FULL CITE], which has been published in final form at https://doi.org/10.1111/ejn.15423. This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Use of Self-Archived Versions. This article may not be enhanced, enriched or otherwise transformed into a derivative work, without express permission from Wiley or by statutory rights under applicable legislation. Copyright notices must not be removed, obscured or modified. The article must be linked to Wiley’s version of record on Wiley Online Library and any embedding, framing or otherwise making available the article or pages thereof by third parties from platforms, services and websites other than Wiley Online Library must be prohibited.
Accepted author manuscript, 1.71 MBLicence: Unspecified

Cite this

Yeung, H. W., Shen, X., Stolicyn, A., De Nooij, L., Harris, M. A., Romaniuk, L., Buchanan, C. R., Waiter, G. D., Sandu, A.-L., McNeil, C. J., Murray, A., Steele, J. D., Campbell, A., Porteous, D., Lawrie, S. M., McIntosh, A. M., Cox, S. R., Smith, K. M., & Whalley, H. C. (2021). Spectral Clustering Based on Structural Magnetic Resonance Imaging and its Relationship with Major Depressive Disorder and Cognitive Ability. European Journal of Neuroscience, 54(6), 6281-6303. https://doi.org/10.1111/ejn.15423

Yeung, HW, Shen, X, Stolicyn, A, De Nooij, L, Harris, MA, Romaniuk, L, Buchanan, CR, Waiter, GD , Sandu, A-L , McNeil, CJ , Murray, A, Steele, JD, Campbell, A, Porteous, D, Lawrie, SM, McIntosh, AM, Cox, SR, Smith, KM & Whalley, HC 2021, 'Spectral Clustering Based on Structural Magnetic Resonance Imaging and its Relationship with Major Depressive Disorder and Cognitive Ability', European Journal of Neuroscience, vol. 54, no. 6, pp. 6281-6303. https://doi.org/10.1111/ejn.15423

@article{e7cbd72068124a1fbb4b3b1f4ca81ab9,

title = "Spectral Clustering Based on Structural Magnetic Resonance Imaging and its Relationship with Major Depressive Disorder and Cognitive Ability",

abstract = "There is increasing interest in using data-driven unsupervised methods to identify structural underpinnings of common mental illnesses, including Major Depressive Disorder (MDD) and associated traits such as cognition. However, studies are often limited to severe clinical cases with small sample sizes and most do not include replication. Here, we examine two relatively large samples with structural magnetic resonance imaging (MRI), measures of lifetime MDD and cognitive variables: Generation Scotland (GS subsample, N = 980), and UK Biobank (UKB, N = 8900); for discovery and replication, using an exploratory approach. Regional measures of FreeSurfer derived cortical thickness (CT), cortical surface area (CSA), cortical volume (CV) and subcortical volume (subCV) were input into a clustering process, controlling for common covariates. The main analysis steps involved constructing participant K-nearest neighbour graphs, and graph partitioning with Markov Stability to determine optimal clustering of participants. Resultant clusters were i) checked whether they were replicated in an independent cohort, and ii) tested for associations with depression status, and cognitive measures. Participants separated into two clusters based on structural brain measurements in GS subsample, with large Cohen's d effect sizes between clusters in higher order cortical regions, commonly associated with executive function and decision making. Clustering was replicated in the UKB sample, with high correlations of cluster effect sizes for CT, CSA, CV and subCV between cohorts across regions. The identified clusters were not significantly different with respect to MDD case-control status in either cohort (GS subsample: pFDR = 0.2239 - 0.6585; UKB: pFDR = 0.2003 - 0.7690). Significant differences in general cognitive ability were, however, found between the clusters for both datasets; for CSA, CV, subCV (GS subsample: d = 0.2529 - 0.3490, pFDR < 0.005; UKB: d = 0.0868 - 0.1070, pFDR < 0.005). Our results suggest that there are replicable natural groupings of participants based on cortical and subcortical brain measures, which may be related to differences in cognitive performance, but not to the MDD case-control status.",

keywords = "Major Depressive Disorder, Cognition, Machine Learning, Clustering, Markov Stability, Structural Neuroimaging",

author = "Yeung, {Hon Wah} and Xueyi Shen and Aleks Stolicyn and {De Nooij}, Laura and Harris, {Mathew A} and Liana Romaniuk and Buchanan, {Colin R} and Waiter, {Gordon D} and Anca-Larisa Sandu and McNeil, {Christopher J} and Alison Murray and Steele, {J Douglas} and Archie Campbell and David Porteous and Lawrie, {Stephen M} and McIntosh, {Andrew M} and Cox, {Simon R} and Smith, {Keith M} and Whalley, {Heather C}",

note = "Acknowledgement This study was supported and funded by the Wellcome Trust Strategic Award “Stratifying Resilience and Depression Longitudinally” (STRADL) (Reference 104036/Z/14/Z), and was also supported by National Institutes of Health (NIH) research grant R01AG054628. Generation Scotland received core support from the Chief Scientist Office of the Scottish Government Health Directorates [CZD/16/6] and the Scottish Funding Council [HR03006] and is currently supported by the Wellcome Trust [216767/Z/19/Z]. The research was conducted using the UKB resource, with approved project number 10279, and the UKB imaging data was processed at the University of Edinburgh Centre for Cognitive Ageing and Cognitive Epidemiology (CCACE) (http://www.ccace.ed.ac.uk), part of the cross-council Lifelong Health and Wellbeing Initiative (MR/K026992/1). CCACE received funding from Biotechnology and Biological Sciences Research Council (BBSRC), Medical Research Council (MRC), and was also supported by Age UK as part of The Disconnected Mind project. Keith M. Smith was supported by Health Data Research UK, an initiative funded by UK Research and Innovation Councils, NIH Research (England) and the UK devolved administrations, and leading medical research charities. Simon R. Cox was supported by Age UK (Disconnected Mind project), the UK Medical Research Council [MRC: MR/R024065/1], and the US National Institutes of Health (NIH) [R01AG054628]. ",

year = "2021",

month = sep,

day = "30",

doi = "10.1111/ejn.15423",

language = "English",

volume = "54",

pages = "6281--6303",

journal = "European Journal of Neuroscience",

issn = "0953-816X",

publisher = "Wiley-Blackwell",

number = "6",

}

TY - JOUR

T1 - Spectral Clustering Based on Structural Magnetic Resonance Imaging and its Relationship with Major Depressive Disorder and Cognitive Ability

AU - Yeung, Hon Wah

AU - Shen, Xueyi

AU - Stolicyn, Aleks

AU - De Nooij, Laura

AU - Harris, Mathew A

AU - Romaniuk, Liana

AU - Buchanan, Colin R

AU - Waiter, Gordon D

AU - Sandu, Anca-Larisa

AU - McNeil, Christopher J

AU - Murray, Alison

AU - Steele, J Douglas

AU - Campbell, Archie

AU - Porteous, David

AU - Lawrie, Stephen M

AU - McIntosh, Andrew M

AU - Cox, Simon R

AU - Smith, Keith M

AU - Whalley, Heather C

N1 - Acknowledgement This study was supported and funded by the Wellcome Trust Strategic Award “Stratifying Resilience and Depression Longitudinally” (STRADL) (Reference 104036/Z/14/Z), and was also supported by National Institutes of Health (NIH) research grant R01AG054628. Generation Scotland received core support from the Chief Scientist Office of the Scottish Government Health Directorates [CZD/16/6] and the Scottish Funding Council [HR03006] and is currently supported by the Wellcome Trust [216767/Z/19/Z]. The research was conducted using the UKB resource, with approved project number 10279, and the UKB imaging data was processed at the University of Edinburgh Centre for Cognitive Ageing and Cognitive Epidemiology (CCACE) (http://www.ccace.ed.ac.uk), part of the cross-council Lifelong Health and Wellbeing Initiative (MR/K026992/1). CCACE received funding from Biotechnology and Biological Sciences Research Council (BBSRC), Medical Research Council (MRC), and was also supported by Age UK as part of The Disconnected Mind project. Keith M. Smith was supported by Health Data Research UK, an initiative funded by UK Research and Innovation Councils, NIH Research (England) and the UK devolved administrations, and leading medical research charities. Simon R. Cox was supported by Age UK (Disconnected Mind project), the UK Medical Research Council [MRC: MR/R024065/1], and the US National Institutes of Health (NIH) [R01AG054628].

PY - 2021/9/30

Y1 - 2021/9/30

N2 - There is increasing interest in using data-driven unsupervised methods to identify structural underpinnings of common mental illnesses, including Major Depressive Disorder (MDD) and associated traits such as cognition. However, studies are often limited to severe clinical cases with small sample sizes and most do not include replication. Here, we examine two relatively large samples with structural magnetic resonance imaging (MRI), measures of lifetime MDD and cognitive variables: Generation Scotland (GS subsample, N = 980), and UK Biobank (UKB, N = 8900); for discovery and replication, using an exploratory approach. Regional measures of FreeSurfer derived cortical thickness (CT), cortical surface area (CSA), cortical volume (CV) and subcortical volume (subCV) were input into a clustering process, controlling for common covariates. The main analysis steps involved constructing participant K-nearest neighbour graphs, and graph partitioning with Markov Stability to determine optimal clustering of participants. Resultant clusters were i) checked whether they were replicated in an independent cohort, and ii) tested for associations with depression status, and cognitive measures. Participants separated into two clusters based on structural brain measurements in GS subsample, with large Cohen's d effect sizes between clusters in higher order cortical regions, commonly associated with executive function and decision making. Clustering was replicated in the UKB sample, with high correlations of cluster effect sizes for CT, CSA, CV and subCV between cohorts across regions. The identified clusters were not significantly different with respect to MDD case-control status in either cohort (GS subsample: pFDR = 0.2239 - 0.6585; UKB: pFDR = 0.2003 - 0.7690). Significant differences in general cognitive ability were, however, found between the clusters for both datasets; for CSA, CV, subCV (GS subsample: d = 0.2529 - 0.3490, pFDR < 0.005; UKB: d = 0.0868 - 0.1070, pFDR < 0.005). Our results suggest that there are replicable natural groupings of participants based on cortical and subcortical brain measures, which may be related to differences in cognitive performance, but not to the MDD case-control status.

AB - There is increasing interest in using data-driven unsupervised methods to identify structural underpinnings of common mental illnesses, including Major Depressive Disorder (MDD) and associated traits such as cognition. However, studies are often limited to severe clinical cases with small sample sizes and most do not include replication. Here, we examine two relatively large samples with structural magnetic resonance imaging (MRI), measures of lifetime MDD and cognitive variables: Generation Scotland (GS subsample, N = 980), and UK Biobank (UKB, N = 8900); for discovery and replication, using an exploratory approach. Regional measures of FreeSurfer derived cortical thickness (CT), cortical surface area (CSA), cortical volume (CV) and subcortical volume (subCV) were input into a clustering process, controlling for common covariates. The main analysis steps involved constructing participant K-nearest neighbour graphs, and graph partitioning with Markov Stability to determine optimal clustering of participants. Resultant clusters were i) checked whether they were replicated in an independent cohort, and ii) tested for associations with depression status, and cognitive measures. Participants separated into two clusters based on structural brain measurements in GS subsample, with large Cohen's d effect sizes between clusters in higher order cortical regions, commonly associated with executive function and decision making. Clustering was replicated in the UKB sample, with high correlations of cluster effect sizes for CT, CSA, CV and subCV between cohorts across regions. The identified clusters were not significantly different with respect to MDD case-control status in either cohort (GS subsample: pFDR = 0.2239 - 0.6585; UKB: pFDR = 0.2003 - 0.7690). Significant differences in general cognitive ability were, however, found between the clusters for both datasets; for CSA, CV, subCV (GS subsample: d = 0.2529 - 0.3490, pFDR < 0.005; UKB: d = 0.0868 - 0.1070, pFDR < 0.005). Our results suggest that there are replicable natural groupings of participants based on cortical and subcortical brain measures, which may be related to differences in cognitive performance, but not to the MDD case-control status.

KW - Major Depressive Disorder

KW - Cognition

KW - Machine Learning

KW - Clustering, Markov Stability

KW - Structural Neuroimaging

UR - http://www.scopus.com/inward/record.url?scp=85114014922&partnerID=8YFLogxK

U2 - 10.1111/ejn.15423

DO - 10.1111/ejn.15423

M3 - Article

C2 - 34390586

SN - 0953-816X

VL - 54

SP - 6281

EP - 6303

JO - European Journal of Neuroscience

JF - European Journal of Neuroscience

IS - 6

ER -

Spectral Clustering Based on Structural Magnetic Resonance Imaging and its Relationship with Major Depressive Disorder and Cognitive Ability

Abstract

Bibliographical note

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Aberdeen Biomedical Imaging Centre

Cite this

Spectral Clustering Based on Structural Magnetic Resonance Imaging and its Relationship with Major Depressive Disorder and Cognitive Ability

Abstract

Bibliographical note

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Equipment

Aberdeen Biomedical Imaging Centre

Cite this