An automatic macaque surface generation pipeline based on structural MRI
Poster No:
1854
Submission Type:
Abstract Submission
Authors:
Yahui Wei1, Haiyan Wang1, Luqi Cheng2, Qi Zhu3, Wen Li1, Congying Chu4, Wim Vanduffel5, Tianzi Jiang4, Lingzhong Fan6
Institutions:
1Brainnetome Center, Institute of Automation, Chinese Academy of Sciences, Beijing, Beijing, 2Guilin University of Electronic Technology, Guilin, Guangxi, 3Cognitive Neuroimaging Unit, INSERM, CEA, Université Paris-Saclay, NeuroSpin Center, Gif/Yvette, Gif/Yvette, 4Institute of Automation, Chinese Academy of Sciences, Beijing, China, 5Department of Neurosciences, Laboratory of Neuro- and Psychophysiology, KU Leuven Medical School, Leuven, Belgium, 6Brainnetome Center, National Laboratory of Pattern Recognition, Institute of Automation, Chinese Aca, Beijing, China
First Author:
Co-Author(s):
Haiyan Wang
Brainnetome Center, Institute of Automation, Chinese Academy of Sciences
Beijing, Beijing
Brainnetome Center, Institute of Automation, Chinese Academy of Sciences
Beijing, Beijing
Qi Zhu
Cognitive Neuroimaging Unit, INSERM, CEA, Université Paris-Saclay, NeuroSpin Center
Gif/Yvette, Gif/Yvette
Cognitive Neuroimaging Unit, INSERM, CEA, Université Paris-Saclay, NeuroSpin Center
Gif/Yvette, Gif/Yvette
Wim Vanduffel
Department of Neurosciences, Laboratory of Neuro- and Psychophysiology, KU Leuven Medical School
Leuven, Belgium
Department of Neurosciences, Laboratory of Neuro- and Psychophysiology, KU Leuven Medical School
Leuven, Belgium
Lingzhong Fan
Brainnetome Center, National Laboratory of Pattern Recognition, Institute of Automation, Chinese Aca
Beijing, China
Brainnetome Center, National Laboratory of Pattern Recognition, Institute of Automation, Chinese Aca
Beijing, China
Introduction:
Macaque monkeys have long been recognized as an essential model organism in neuroscience due to their close evolutionary relationship with humans. Constructing the brain surface of macaques from structural MRI data is fundamental for multi-modal MRI data analysis, enabling the extraction of surface-related indices and the mapping of other modalities onto the surface for better visualization and analysis. However, existing surface reconstruction methods are primarily designed for the human brain, leading to extensive manual adjustments and suboptimal performance when applied to macaques. In this work, we address these challenges by optimizing multiple steps specifically for macaque brains and leveraging state-of-the-art macaque brain segmentation tools, macaque-specific template, and a detailed macaque brain atlas. This surface reconstruction pipeline, tailored for macaques, is user-friendly and demonstrates high performance across datasets of varying quality.
Methods:
The whole pipeline can be summarized as the following steps:
(1) Skull stripping and brain segmentation on each run based on the deep learning-based tool, i.e. nBEST (Zhong et al., 2024)
(2) Automatically choose high-quality runs based on MRIQC (Esteban et al., 2017), in which the brain segmentation results will also be taken into consideration.
(3) If more than one runs exist, get the average MRI after rigid registration.
(4) Automatically orientation correction based on linear registration.
(5) Conform of data without interpolation to meet FreeSurfer 's request.
(6) Register individual T1 to a high-quality MEBRAIN macaque template (Balan et al., 2024) to get the atlas label from Macaque BNA (Lu et al., 2024).
(7) Tissue segmentation on the average MRI using nBEST.
(8) A newly developed tissue-based method to improve contrast between different tissues.
(9) The tissue-guided N4 bias field correction (Tustison et al., 2010) was employed to mitigate the increased intra-tissue bias caused by contrast improvement.
(10) Prepare the image file for generating surfaces based on brain segmentation and atlas label.
(11-12) Using FreeSurfer to generate white surface and pial surface using the best parameters. If T2 is available, T2 will be used to optimize the surface generation.
(1) Skull stripping and brain segmentation on each run based on the deep learning-based tool, i.e. nBEST (Zhong et al., 2024)
(2) Automatically choose high-quality runs based on MRIQC (Esteban et al., 2017), in which the brain segmentation results will also be taken into consideration.
(3) If more than one runs exist, get the average MRI after rigid registration.
(4) Automatically orientation correction based on linear registration.
(5) Conform of data without interpolation to meet FreeSurfer 's request.
(6) Register individual T1 to a high-quality MEBRAIN macaque template (Balan et al., 2024) to get the atlas label from Macaque BNA (Lu et al., 2024).
(7) Tissue segmentation on the average MRI using nBEST.
(8) A newly developed tissue-based method to improve contrast between different tissues.
(9) The tissue-guided N4 bias field correction (Tustison et al., 2010) was employed to mitigate the increased intra-tissue bias caused by contrast improvement.
(10) Prepare the image file for generating surfaces based on brain segmentation and atlas label.
(11-12) Using FreeSurfer to generate white surface and pial surface using the best parameters. If T2 is available, T2 will be used to optimize the surface generation.
·Overall Workflow
Results:
We first show the performance of our proposed tissue-guided method to improve data quality. As illustrated in Figure 2A, the final corrected image shows higher brain contrast than raw images, which ensures the accuracy of following surface generation. For testing the whole pipeline's performance on surface generation, we tested it on high-quality, media-quality, and low-quality macaque structural MRI datasets, and find that our pipeline shows good performance on high- and media-quality data, and acceptable performance on low-quality data (Figure 2B).
·Results of Augmentation and Surface Reconstruction
Conclusions:
In conclusion, we developed and integrated several state-of-the-art methods into a comprehensive pipeline for macaque cortical surface reconstruction. To enhance accessibility and facilitate further research, we plan to make this tool openly available, which will facilitate cross-species comparative studies and provide deeper insights into brain evolution.
Modeling and Analysis Methods:
Image Registration and Computational Anatomy
Methods Development 2
Segmentation and Parcellation
Neuroinformatics and Data Sharing:
Brain Atlases
Workflows 1
Keywords:
Data Registration
Machine Learning
MRI
Open-Source Code
Spatial Normalization
STRUCTURAL MRI
Workflows
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
2. Balan, P. F., Zhu, Q., Li, X., Niu, M., Rapan, L., Funck, T., . . . Vanduffel, W. (2024). MEBRAINS 1.0: A new population-based macaque atlas. Imaging Neuroscience, 2, 1-26. doi:10.1162/imag_a_00077
3. Esteban, O., Birman, D., Schaer, M., Koyejo, O. O., Poldrack, R. A., & Gorgolewski, K. J. (2017). MRIQC: Advancing the automatic prediction of image quality in MRI from unseen sites. PLoS One, 12(9), e0184661. doi:10.1371/journal.pone.0184661
4. Fischl, B. (2012). FreeSurfer. Neuroimage, 62(2), 774-781. doi:10.1016/j.neuroimage.2012.01.021
5. Lu, Y., Cui, Y., Cao, L., Dong, Z., Cheng, L., Wu, W., . . . Jiang, T. (2024). Macaque Brainnetome Atlas: A multifaceted brain map with parcellation, connection, and histology. Sci Bull (Beijing), 69(14), 2241-2259. doi:10.1016/j.scib.2024.03.031
6. Tustison, N. J., Avants, B. B., Cook, P. A., Zheng, Y., Egan, A., Yushkevich, P. A., & Gee, J. C. (2010). N4ITK: improved N3 bias correction. IEEE Trans Med Imaging, 29(6), 1310-1320. doi:10.1109/TMI.2010.2046908
7. Zhong, T., Wu, X., Liang, S., Ning, Z., Wang, L., Niu, Y., . . . Zhang, Y. (2024). nBEST: Deep-learning-based non-human primates Brain Extraction and Segmentation Toolbox across ages, sites and species. Neuroimage, 295, 120652. doi:10.1016/j.neuroimage.2024.120652