Mapping Terra Incognita: Exploring Subcortical Structures in Major Depressive Disorder with 7T MRI

2165 

Abstract Submission 

Dilan Yucel¹, Jurjen Heij², Wietske van der Zwaag³, Matthan Caan⁴, Anneke Alkemade⁵, Pierre-Louis Bazin⁶, Birte Forstmann⁷, Moji Aghajani⁸, Guido Wingen⁹

¹Amsterdam UMC, Amsterdam, Noord Holland, ²Spinoza Centre for Neuroimaging, KNAW, AMSTERDAM, North Netherlands, ³Spinoza Centre for Neuroimaging, KNAW, Amsrwedam, North Netherlands, ⁴Amsterdam Neuroscience, Amsterdam, North Netherlands, ⁵University of Amsterdam, Amsterdam, Netherlands, ⁶Full brain picture Analytics, Leiden, Zuid Holland, ⁷Integrative Model-based Cognitive Neuroscience Research Unit, Amsterdam, Netherlands, ⁸Amsterdam UMC location Vrije Universiteit Amsterdam, Amsterdam, North Netherlands, ⁹Amsterdam UMC location University of Amsterdam, Amsterdam, Noord-Holland

Dilan Yucel  
Amsterdam UMC
Amsterdam, Noord Holland

Jurjen Heij  
Spinoza Centre for Neuroimaging, KNAW
AMSTERDAM, North Netherlands

Wietske van der Zwaag  
Spinoza Centre for Neuroimaging, KNAW
Amsrwedam, North Netherlands

Matthan Caan  
Amsterdam Neuroscience
Amsterdam, North Netherlands

Anneke Alkemade  
University of Amsterdam
Amsterdam, Netherlands

Pierre-Louis Bazin  
Full brain picture Analytics
Leiden, Zuid Holland

Birte Forstmann  
Integrative Model-based Cognitive Neuroscience Research Unit
Amsterdam, Netherlands

Moji Aghajani  
Amsterdam UMC location Vrije Universiteit Amsterdam
Amsterdam, North Netherlands

Guido Wingen  
Amsterdam UMC location University of Amsterdam
Amsterdam, Noord-Holland

Despite significant advances in recent years, our understanding of the neuropathological mechanisms underlying major depressive disorder (MDD) remains limited. This restriction is particularly evident in subcortical regions, where neuroimaging techniques at lower magnetic field strengths encounter challenges related to resolution and detail. Animal and theoretical models of affective psychopathology propose the involvement of specific subcortical regions and their subnuclei [4]. Moreover, recent research has highlighted the importance of detailed subcortical investigations in MDD [3]. However, current neuroimaging techniques lack the necessary precision to thoroughly investigate the structural integrity of these areas. Consequently, our understanding of subcortical brain systems in the context of MDD is constrained, significantly hindering the practical application and translation of research findings in this field. Nevertheless, the introduction of ultra-high-resolution MRI protocols, specifically those acquiring data at 7 Tesla (7T), holds significant potential in addressing these limitations and facilitating a more comprehensive understanding of the neurobiological foundations of MDD [1]. Aligned with this objective, we conducted a cutting-edge 7T imaging study to unravel the pathophysiology of MDD by investigating subcortical structures.

We acquired T1-weighted MP2RAGE images and T2-weighted quantitative images at ultra high-field (7T MRI) from a cohort of 71 individuals, including 57 MDD patients and 14 healthy controls. The segmentation of subcortical regions was performed using Nighres, a novel open-source parcellation algorithm designed for processing high-resolution neuroimaging data [2]. Thirty-one subcortical structures were sampled onto each subject's volumetric space, and the T1-/T2*-data corresponding to these structures were processed using Nighres' profile sampling module. Statistical analysis included Bayesian ANCOVA and Bayesian correlation analysis implemented in JASP, aiming to elucidate effects observed across groups and within the MDD patient group. Segmentations of subcortical regions, lateral ventricles, and total volume were compared between patients and controls using Bayesian ANCOVA, controlling for age and sex. We also explored the effects of symptom severity, medication (antidepressants/psychotropics), childhood trauma, and comorbid anxiety using a between-patients dimensional approach.

Both Bayesian ANCOVA and regression analyses yielded anecdotal to moderate evidence in favor of the absence of a group effect of major depressive disorder across all thirty-one subcortical regions (1<Bayes Factor (BF)<10). Bayesian ANCOVA demonstrated anecdotal to moderate evidence for the effects of medication, childhood trauma, and comorbid anxiety on major depressive disorder across all thirty-one subcortical subnuclei (1<Bayes Factor (BF)<10). Bayesian regression identified no association between the symptom severity of major depressive disorder and subcortical brain volumes (1<Bayes Factor (BF)<10). Additionally, sample characteristics such as mean age, gender, and medication use did not moderate the alterations in subcortical volumes, and the T1-/T2* metrics.

Our study stands as one of the scarce pioneering efforts delving into the exploration of subcortical nuclei in major depressive disorder through the application of ultra-high field imaging. Despite the Bayesian framework offering evidence of weak strengths in identifying depression biomarkers in our study, the utilization of ultra-high field imaging persists as a compelling avenue for unraveling the neuropathology of Major Depressive Disorder (MDD). A future direction involving extensive mapping of the subcortex using ultra-high field imaging holds the promise of yielding profound insights into the potential role of these subcortical regions as prominent biomarkers in the neurobiological modeling of MDD.

Psychiatric (eg. Depression, Anxiety, Schizophrenia)

Bayesian Modeling

Segmentation and Parcellation ²

Cortical Anatomy and Brain Mapping

Subcortical Structures ¹

Affective Disorders

Computational Neuroscience

Data analysis

HIGH FIELD MR

MRI

Psychiatric Disorders

Segmentation

STRUCTURAL MRI

Sub-Cortical