Facilitating seed searching for resting state fMRI data analysis in large-scale study
Poster No:
1478
Submission Type:
Abstract Submission
Authors:
Mingyi Li1, Ajay Nemani1, Katherine Koenig1, Yugo Boureima1, Christopher MacClugage1, Jessica Padilla1, Mark Lowe1
Institutions:
1Cleveland Clinic, Cleveland, OH
First Author:
Co-Author(s):
Introduction:
When using seed-based methods to analyze brain resting state fMRI (rs-fMRI) data, the selection of the seed has large influence on the results. We developed a machine learning based method to automatically produce prospective seeds by utilizing anatomical MRI and rs-fMRI data. Shown in our previous study, one automatically generated seed could closely match manually picked seed in 13 out of 19 subjects[1]. The method was not fully automatic because human manual review was demanded to pick the right seed, if there was one, out of 10~20 automatically generated seed candidates. Starting from year 2022, Cleveland Clinic Foundation launched a large scale brain study(CCBS) including collecting anatomical and rs-fMRI data from thousands of healthy subjects projected in 20 years[2]. Thus, more automatic processing and less manual operation is highly preferred in data analysis. We tuned our previously developed method to meet such challenge by reducing the amount of automatically generated seed candidates to less than five. Testing results on fifty healthy subjects in CCBS study will be shown in this abstract.
Methods:
Data collection: 50 cognitively healthy subjects (age=64.2+/-7.0, 31 female) were selected from the IRB approved CCBS study. Whole brain functional imaging was performed on a 3T scanner using echo planar imaging (104x104x51matrix, 2.5mm isotropic voxel size, 260 volumes). High resolution brain T1w images (208 x240x256 matrix , 1mm isotropic voxel size)were collected for anatomical context along with dual echo field maps.
Data processing: Each rs-fMRI dataset was corrected for motion, physiologically-based nuisances, and B0 distortions. It was then low-pass filtered and spatially filtered. Brain T1 image was parcellated into ROIs by using FreeSurfer[3] and then registered to rs-fMRI image space by using "bbregister" tool[4]. The ROI covering the left/right PCC was used as seed searching region for automatic seed generation method (Left side and right side are processed separately). From rs-fMRI data, the global connectivity between each voxel in the seed searching ROI and all other brain cortex voxels were computed and then the connectivity distribution was fitted into a Gaussian distribution [5]. The feature vector was formed by counting the number of voxels whose connectivity value was outside three standard deviations, in the parcellated cortex ROIs. The above feature vector forming step is shown in figure 1. Then the feature vectors of all the voxels in the searching region was feed into a size 4x4 SOM classifier in Matlab. SOM cluster was further divided into spatially aggregated sub-clusters by using k-means clustering. At last, voxels with weak correlation to other voxels in the same cluster were discarded. The remaining voxels formed the seed clusters.
Using each subject's rs-fMRI data, seeds in PCC were also manually located by experts through the "instacorr" method in AFNI[6].
Seeds acquired through the automatic method were compared to those picked through the manual method.
Data processing: Each rs-fMRI dataset was corrected for motion, physiologically-based nuisances, and B0 distortions. It was then low-pass filtered and spatially filtered. Brain T1 image was parcellated into ROIs by using FreeSurfer[3] and then registered to rs-fMRI image space by using "bbregister" tool[4]. The ROI covering the left/right PCC was used as seed searching region for automatic seed generation method (Left side and right side are processed separately). From rs-fMRI data, the global connectivity between each voxel in the seed searching ROI and all other brain cortex voxels were computed and then the connectivity distribution was fitted into a Gaussian distribution [5]. The feature vector was formed by counting the number of voxels whose connectivity value was outside three standard deviations, in the parcellated cortex ROIs. The above feature vector forming step is shown in figure 1. Then the feature vectors of all the voxels in the searching region was feed into a size 4x4 SOM classifier in Matlab. SOM cluster was further divided into spatially aggregated sub-clusters by using k-means clustering. At last, voxels with weak correlation to other voxels in the same cluster were discarded. The remaining voxels formed the seed clusters.
Using each subject's rs-fMRI data, seeds in PCC were also manually located by experts through the "instacorr" method in AFNI[6].
Seeds acquired through the automatic method were compared to those picked through the manual method.
·Figure1. Generating feature vector by combining rs-fMRI connectivity and T1 parcellation. Panel A: Z-map, Panel B: Z-score distribution, Panel C: Freesurfer parcellation, Panel D: feature vectors.
Results:
In 17 out of 50 subjects, there was one automatically generated seed overlapping with manually picked seed in some degree. The percentage of matched seeds drop substantially comparing to previous results with 10~20 seed candidates.
One typical matched case is shown in figure 2. Please note that the automatically generated seed followed the anatomical structure in much better way than the manually picked and grown seed.
One typical matched case is shown in figure 2. Please note that the automatically generated seed followed the anatomical structure in much better way than the manually picked and grown seed.
·Figure 2. Matched seeds. Top row shows automatically generated seed and bottom row shows manually picked seed.
Conclusions:
Although it will be very helpful to apply the method with less than five seed candidates on large scale dataset to save human efforts, the downgrade in performance was quite high. So, it is worth the effort to test the method with 10~20 seed candidates on the same dataset. Hopefully, there exist common features shared by the matched seeds across subjects, so that seed candidates with similar features as the matched seeds can be prioritized for manual review. In this way, it is equal to reducing the seed candidates.
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural)
fMRI Connectivity and Network Modeling 1
Methods Development 2
Keywords:
Data analysis
FUNCTIONAL MRI
Machine Learning
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?
Are you Internal Review Board (IRB) certified? Please note: Failure to have IRB, if applicable will lead to automatic rejection of abstract.
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] Cleveland Clinic Foundation. (2022). Cleveland Clinic Brain Study. https://my.clevelandclinic.org/departments/neurogical/research-innovations/brain-study
[3] Fischl B. (2012), FreeSurfer, NeuroImage, 2012, 62(2): 774-781.
[4] DN Greve, et al. Accurate and robust brain image alignment using boundary-basedregistration, Neuroimage. 2009; 48:63-72.
[5] MJ Lowe, et al. Functional Connectivity in Single and Multislice Echoplanar Imaging Using Resting-State Fluctuations, NeuroImage. 1998; 7(1):119-132.
[6] Cox RW. AFNI: What a long strange trip it’s been, NeuroImage. 2012; 62(2):743-747.