Shifts in structural connectome manifolds in patients with migraine

1504 

Abstract Submission 

Eunchan Noh¹, YEONGJUN PARK², Yurim Jang³, Jong Young Namgung⁴, Mi Ji Lee^5,6, Bo-yong Park^4,7,8

¹College of Medicine, Inha University, Incheon, Republic of Korea, ²Department of Computer Engineering, Inha University, Incheon, Republic of Korea, ³Artificial Intelligence Convergence Research Center, Inha University, Incheon, Republic of Korea, ⁴Department of Data Science, Inha University, Incheon, Republic of Korea, ⁵Department of Neurology, Seoul National University Hospital, Seoul, Republic of Korea, ⁶Department of Neurology, Seoul National University College of Medicine, Seoul, Republic of Korea, ⁷Department of Statistics and Data Science, Inha University, Incheon, Republic of Korea, ⁸Center for Neuroscience Imaging Research, Institute for Basic Science, Suwon, Republic of Korea

Eunchan Noh  
College of Medicine, Inha University
Incheon, Republic of Korea

YEONGJUN PARK  
Department of Computer Engineering, Inha University
Incheon, Republic of Korea

Yurim Jang  
Artificial Intelligence Convergence Research Center, Inha University
Incheon, Republic of Korea

Jong Young Namgung  
Department of Data Science, Inha University
Incheon, Republic of Korea

Mi Ji Lee  
Department of Neurology, Seoul National University Hospital|Department of Neurology, Seoul National University College of Medicine
Seoul, Republic of Korea|Seoul, Republic of Korea

Bo-yong Park  
Department of Data Science, Inha University|Department of Statistics and Data Science, Inha University|Center for Neuroscience Imaging Research, Institute for Basic Science
Incheon, Republic of Korea|Incheon, Republic of Korea|Suwon, Republic of Korea

Migraine is a neurological condition showing a phase of headache often accompanied by vomiting and sensitivity to sound or light [1]. Prior research based on functional magnetic resonance imaging (fMRI) found alterations in functional connectivity in patients with migraine [2][3], but structural connectome alterations have not been fully investigated. In this study, we aimed to identify alterations in structural connectome organization in patients with migraine using manifold learning techniques that project high-dimensional data onto the low-dimensional manifold space.

We obtained T1-weighted and diffusion MRI of 47 migraine patients (age = 34.3 ± 8.3, 74.5% female) and 41 healthy controls (age = 35.2 ± 7.7, 75.6% female). The T1-weighted data were preprocessed using a FuNP surface-based pipeline [4], and diffusion MRI data were preprocessed using MRtrix3 [5]. The neuronal streamlines were estimated using probabilistic tractography, and the structural connectivity matrix was constructed based on the sub-parcellation of the Desikan-Killiany atlas divided into 300 regions [6]. Low-dimensional eigenvectors were estimated by a nonlinear dimensionality reduction technique - diffusion map embedding. Individual eigenvectors were aligned to a template eigenvector, which was generated from the group representative matrix computed using distance-dependent thresholding [7]. We defined the manifold eccentricity feature by calculating the Euclidean distance between each data point and the center of the data on the manifold space [8] (Fig. 1). The differences in the manifold eccentricity between patients with migraine and healthy controls were estimated using an independent samples t-test while controlling for age and sex. The multiple comparisons across brain regions were corrected using a false discovery rate (FDR) < 0.05. The between-group differences in subcortical connectivity were assessed using the degree values of the subcortico-cortical connectivity.

We found significant shifts in the manifold eccentricity in the orbitofrontal cortex, temporal pole, and somatomotor regions. The largest effect was observed in the limbic areas when we summarized the statistics according to seven functional networks and four cortical hierarchical levels (Fig. 2A). Additionally, significant shifts in degree values of the subcortico-cortical connectivity were observed in the caudate, amygdala, thalamus, and accumbens (Fig. 2B).

Leveraging manifold learning techniques, we found that patients with migraine show shifts in structural connectome organization, particularly in the limbic and amygdala-caudate systems. Our findings may provide a clue for understanding the large-scale brain organization in migraine.

Funds:
This research was supported by the National Research Foundation of Korea (NRF2020R1A2B5B01001826; NRF-2021R1F1A1052303; NRF2022R1A5A7033499), Institute for Information and Communications Technology Planning and Evaluation (IITP) funded by the Korea Government (MSIT) (No. 2022-0-00448, Deep Total Recall: Continuous Learning for Human-Like Recall of Artificial Neural Networks; No. RS-2022-, Artificial Intelligence Convergence Innovation Human Resources Development. 2022-0-00448, Deep Total Recall: Continuous Learning for Human-Like Recall of Artificial Neural Networks; No. RS-2022-00155915, Artificial Intelligence Convergence Innovation Human Resources Development (Inha University); No. RS-2022R1A5A7033499, Institute for Information and Communications Technology Planning and Evaluation (IITP) funded by the Korean Government (MSIT) (No. RS-2022R1A5A7033499) 021-0-02068, Artificial Intelligence Innovation Hub), Institute for Basic Science (IBSR015-D1), and the Korea Medical Device Development Fund grant funded by the Korean government (the Ministry of Science and ICT, the Ministry of Trade, Industry and Energy, the Ministry of Health & Welfare, the Ministry of Food and Drug Safety) (Project number: RS-2023-00229484).

Classification and Predictive Modeling

Connectivity (eg. functional, effective, structural) ¹

Diffusion MRI Modeling and Analysis ²

Methods Development

Univariate Modeling

Data analysis

Headache

Limbic Systems

Multivariate

Pain

Sub-Cortical

Other - Diffusion MRI