The impact of COVID-19 on the integrity of deep and superficial white matter bundles
Poster No:
2201
Submission Type:
Abstract Submission
Authors:
Claudio Roman1, Patricio Carvajal2, Patricia Soto3, Leonie Kausel3, Ximena Stecher4, Alejandra Figueroa-Vargas5, Francisco Zamorano6, María Paz Martinez-Molina3, Gabriela Valdebenito3, Julio Marquez3, Mauricio Aspé-Sánchez3, Rodrigo Henríquez-Ch7, Francisco Aboitiz8, Paula Muñoz-Venturelli9, Carla Manterola10, Reinaldo Uribe11, Czischke Karen11, WAEL EL-DEREDY1, Pamela Guevara12, Pablo Billeke13
Institutions:
1Universidad de Valparaíso, Valparaíso, Valparaíso, 2University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS), Santiago, Región Metropolitana, 3University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS), Santiago de Chile, Metropolitana, 4Neuroradiología, Departamento de Radiología, Clínica Alemana de Santiago, Santiago de Chile, Metropolitana, 5University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS), Santiago, Santiago, 6Universidad San Sebastián, Santiago de Chile, Metropolitana, 7Pontificia Universidad Católica de Chile, Centro Interdisciplinario de Neurociencias, Santiago de Chile, Metropolitana, 8Pontificia Universidad Católica de Chile - Centro Interdisciplinario de Neurociencias, Santiago de Chile, Metropolitana, 9Centro de Estudios Clínicos, Facultad de Medicina, Universidad del Desarrollo, Santiago de Chile, Metropolitana, 10Universidad de Chile, Santiago de Chile, Metropolitana, 11Pontificia Universidad Católica de Chile, Santiago de Chile, Metropolitana, 12Universidad de Concepción, Concepcion, Región del Bio-Bio, 13University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS), Santiago, Chile
First Author:
Co-Author(s):
Patricio Carvajal
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago, Región Metropolitana
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago, Región Metropolitana
Patricia Soto
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
Leonie Kausel
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
Ximena Stecher
Neuroradiología, Departamento de Radiología, Clínica Alemana de Santiago
Santiago de Chile, Metropolitana
Neuroradiología, Departamento de Radiología, Clínica Alemana de Santiago
Santiago de Chile, Metropolitana
Alejandra Figueroa-Vargas
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago, Santiago
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago, Santiago
María Paz Martinez-Molina
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
Gabriela Valdebenito
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
Julio Marquez
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
Mauricio Aspé-Sánchez
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago de Chile, Metropolitana
Rodrigo Henríquez-Ch
Pontificia Universidad Católica de Chile, Centro Interdisciplinario de Neurociencias
Santiago de Chile, Metropolitana
Pontificia Universidad Católica de Chile, Centro Interdisciplinario de Neurociencias
Santiago de Chile, Metropolitana
Francisco Aboitiz
Pontificia Universidad Católica de Chile - Centro Interdisciplinario de Neurociencias
Santiago de Chile, Metropolitana
Pontificia Universidad Católica de Chile - Centro Interdisciplinario de Neurociencias
Santiago de Chile, Metropolitana
Paula Muñoz-Venturelli
Centro de Estudios Clínicos, Facultad de Medicina, Universidad del Desarrollo
Santiago de Chile, Metropolitana
Centro de Estudios Clínicos, Facultad de Medicina, Universidad del Desarrollo
Santiago de Chile, Metropolitana
Pablo Billeke
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago, Chile
University of Desarrollo UDD · Centro de Investigación en Complejidad Social (CICS)
Santiago, Chile
Introduction:
The recent coronavirus disease (COVID-19) has significantly impacted public health, affecting not only the respiratory system but also other organs such as the brain [Spudich et al., 2022].
Studies have shown that recovered patients may experience neuropsychiatric alterations and thinning of certain cerebral cortex areas, especially those connected to the primary olfactory cortex. It has also been suggested that patients who have anosmia could present affection of orbitofrontal regions [Douaud et al., 2022].
This study aims to detect changes in the integrity of specific fiber bundles using two brain fiber atlases: the first composed of known deep white matter (DWM) bundles and the second composed of short association bundles of the superficial white matter (SWM).
Studies have shown that recovered patients may experience neuropsychiatric alterations and thinning of certain cerebral cortex areas, especially those connected to the primary olfactory cortex. It has also been suggested that patients who have anosmia could present affection of orbitofrontal regions [Douaud et al., 2022].
This study aims to detect changes in the integrity of specific fiber bundles using two brain fiber atlases: the first composed of known deep white matter (DWM) bundles and the second composed of short association bundles of the superficial white matter (SWM).
Methods:
A total of 101 subjects participated in this study: 72 patients recovered from COVID-19 and 29 healthy controls.
Images were acquired with a 3T Siemens Skyra scanner. A diffusion-weighted scan sequence (DTI) was acquired (voxel size: 1.8x1.8x2.4 mm, slices: 64, b-value: 1000s/mm2, directions: 70, FoV: 240mm, TR = 10.0s, TE = 95 ms).
Fig. 1 shows a general outline of the processing pipeline to identify bundles with significant differences between groups.
The DTI model was computed using DSI Studio software (http://dsi-studio.labsolver.org/), which extracted diffusion measures such as fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD).
Deterministic tractography algorithm [Yeh et al., 2013] was applied using the following parameters: angular threshold=60◦, step size= 1mm, minimum length= 30mm, maximum length= 250mm, smoothing= 0.5, and QA threshold=0.
The fiber segmentation was performed using both a DWM bundle atlas [Guevara et al., 2012], which consists of 36 known bundles, and a SWM bundle atlas [Román et al., 2022], composed of 525 short bundles, from which the 209 most stable bundles for deterministic tractography were selected.
An automatic segmentation algorithm based on the maximum Euclidean distance between the corresponding points of two fibers was used [Vázquez et al., 2019].
For each segmented bundle, a binary mask was computed to calculate the average of each diffusion measure. Then, a t-test was computed to compare the patient and control groups, where statistical significance was considered when p-value<0.05.
Images were acquired with a 3T Siemens Skyra scanner. A diffusion-weighted scan sequence (DTI) was acquired (voxel size: 1.8x1.8x2.4 mm, slices: 64, b-value: 1000s/mm2, directions: 70, FoV: 240mm, TR = 10.0s, TE = 95 ms).
Fig. 1 shows a general outline of the processing pipeline to identify bundles with significant differences between groups.
The DTI model was computed using DSI Studio software (http://dsi-studio.labsolver.org/), which extracted diffusion measures such as fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD).
Deterministic tractography algorithm [Yeh et al., 2013] was applied using the following parameters: angular threshold=60◦, step size= 1mm, minimum length= 30mm, maximum length= 250mm, smoothing= 0.5, and QA threshold=0.
The fiber segmentation was performed using both a DWM bundle atlas [Guevara et al., 2012], which consists of 36 known bundles, and a SWM bundle atlas [Román et al., 2022], composed of 525 short bundles, from which the 209 most stable bundles for deterministic tractography were selected.
An automatic segmentation algorithm based on the maximum Euclidean distance between the corresponding points of two fibers was used [Vázquez et al., 2019].
For each segmented bundle, a binary mask was computed to calculate the average of each diffusion measure. Then, a t-test was computed to compare the patient and control groups, where statistical significance was considered when p-value<0.05.
·Fig. 1. Schematic illustrating the process to identify bundles with significant differences between two groups.
Results:
For the fiber segmentation, we used the predefined thresholds, between 6 and 8mm for SWM bundles and between 10 and 20mm for DWM bundles.
After applying the t-test, 3 bundles with significant differences were found in some of the diffusion measures for the DWM atlas, and 29 bundles for the SWM atlas.
Fig. 2 shows the bundles with significant differences between COVID-19 recovered subjects and controls.
The AD is the diffusion measure that showed significant differences between groups in a greater number of bundles (DWM: 3 bundles, SWM: 21 bundles). To complement the quantification of differences between groups, the effect size was calculated using Cohen's d. For the bundles with significant differences, an average Cohen's d of 0.64 was obtained, while the average Cohen's d considering all the bundles was 0.27.
After applying the t-test, 3 bundles with significant differences were found in some of the diffusion measures for the DWM atlas, and 29 bundles for the SWM atlas.
Fig. 2 shows the bundles with significant differences between COVID-19 recovered subjects and controls.
The AD is the diffusion measure that showed significant differences between groups in a greater number of bundles (DWM: 3 bundles, SWM: 21 bundles). To complement the quantification of differences between groups, the effect size was calculated using Cohen's d. For the bundles with significant differences, an average Cohen's d of 0.64 was obtained, while the average Cohen's d considering all the bundles was 0.27.
·Fig. 2. Bundles with significant differences between diffusion metrics of COVID-19 recovered subjects and controls who have not had the disease. First row: DWM bundles. Second row: SWM bundles.
Conclusions:
In this study, the difference in bundle integrity between COVID-19 recovered subjects and controls who have not had the disease was analyzed. For this, metrics obtained from DTI in bundles of the DWM and SWM were compared. Among the known bundles of the DWM with significant differences are the fornix, thalamic radiations, and the corticospinal tract of the right hemisphere. In the case of the SWM bundles, the region that presented the highest number of bundles is the precentral region with 10 bundles. The bundles with significant differences between groups also showed a higher Cohen's d compared to the other bundles.
We aim to continue this study complementing it with functional magnetic resonance data to determine both structural and functional consequences due to the disease.
We aim to continue this study complementing it with functional magnetic resonance data to determine both structural and functional consequences due to the disease.
Modeling and Analysis Methods:
Diffusion MRI Modeling and Analysis 2
Segmentation and Parcellation
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
White Matter Anatomy, Fiber Pathways and Connectivity 1
Keywords:
Segmentation
Tractography
White Matter
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
1|2Indicates the priority used for review
Provide references using author date format
Yeh, F.-C., (2013), ‘Deterministic diffusion fiber tracking improved by quantitative anisotropy,’ PloS one, vol. 8, no. 11, pp. e80713.
Guevara, P. (2012), ‘Automatic fiber bundle segmentation in massive tractography datasets using a multi-subject bundle atlas,’ NeuroImage, vol. 61, no. 4, pp. 1083–1099.
Roman, C. (2022), ‘Superficial white matter bundle atlas based on hierarchical fiber clustering over probabilistic tractography data,’ NeuroImage, vol. 262, pp. 119550.
Spudich, S. (2022), ‘Nervous system consequences of COVID-19’. Science, vol. 375, no. 6578, pp. 267-269.
Vazquez, A. (2019), ‘Parallel optimization of fiber bundle segmentation for massive tractography datasets,” in 2019 IEEE 16th International Symposium on Biomedical Imaging.
Acknowledgements: This work has received funding by ANID, Chile: FONDECYT Postdoctorado 3220729, FONDECYT 1221665, ANILLO ACT210053, Basal FB0008 (AC3E) and Basal FB210017 (CENIA).