Functional and structural connectome estimates predict post-stroke motor recovery from lesions only
Poster No:
1693
Submission Type:
Abstract Submission
Authors:
Christie Gillies1, Senha Pandya1, Keith Jamison1, Aaron Boes2, Amy Kuceyeski3
Institutions:
1Weill Cornell Medicine, New York, NY, 2University of Iowa, Iowa City, IA, 3Cornell, Ithaca, NY
First Author:
Co-Author(s):
Introduction:
Stroke is a leading cause of long-term disability, yet the neural basis of brain recovery remains unclear, hindering brain-based therapeutics. Advanced imaging, like diffusion and resting-state fMRI, has been used to measure structural (SC) and functional connectomes (FC) in stroke patients but faces challenges: patient discomfort, limited resources, lack of required expertise to process such images, and increased noise in these modalities due to stroke pathology. Most stroke patients, however, undergo clinical MRI, which delineates the stroke lesion.This project leverages two lab-developed tools to estimate SC (eSC) and FC (eFC) directly from lesion masks. The NeMo Tool is a database of tractography results from healthy controls that will remove streamlines passing through a lesion mask to provide eSC (Kuceyeski et al., 2013). The Krakencoder, a joint autoencoder framework that maps between SC and FC, and can be used to estimate FC from the NeMo Tool's estimates of SC (Jamison et al., 2024). These eSCs and eFCs offer a novel way to capture the impact of lesions on brain network dynamics without acquiring advanced MRI in patient populations. Here, we predict motor scores from eSC and eFC and compare the accuracy to predictions from observed FC (oFC). If we can achieve good prediction accuracy with estimated connectome measures, this approach has the potential to inform personalized recovery strategies in an efficient way.
Methods:
The data used in this project is from a cohort of 132 adults aged 18–71 (45% female) who had strokes and were admitted to the School of Medicine at Washington University. The data includes MRI and a battery of neuropsychological tests. Scanning was performed with a Siemens 3T Tim-Trio scanner and included resting-state fMRI (TR = 2000ms, 4mm voxels, 6-8 scans for 30 mins total) (Siegel et al., 2016). Lesion masks were created in a semi-automated way and hand-checked by a clinician for validity. Lesion masks were then fed into the NeMo Tool to obtain eSC and these outputs then fed into the Krakencoder to obtain eFC. Functional MRI was processed using standard pipelines and an oFC was obtained for each stroke subject. First, we wanted to verify that the eFC matched somewhat to the oFC, so we calculated their identifiability - or how well an individual's eFC matched their oFC (compared to other individuals' oFCs). To calculate identifiability, a matrix was created containing the correlations of all subjects' vectorized eFC compared to all subjects' oFCs. The percentile of the descending-order rank of the diagonal value (compared to the rest of the row's values) was found, as this is how well an individual's eFC matched their oFC. Ridge regression was used to predict motor score recovery from eSC, eFC and oFC.
Results:
The NeMo+Krakencoder's eFC, derived only from stroke patients' lesion masks, had identifiability (compared to oFC) of 59.8% which was significantly higher than chance levels (permutation-based p<0.003), (Fig. 1). The NeMo Tool's eSC and the NeMo+Krakencoder-derived eFC resulted in significantly more accurate predictions of motor scores at baseline compared to oFC (eSC pFDR<1e-3 and eFC pFDR = 1e-3) derived from fMRI data in the stroke patients themselves (Fig. 2). These results highlight the NeMo and Krakencoder's ability to capture behaviorally meaningful structural and functional brain network disruptions due to stroke.
·Figure 1
·Figure 2
Conclusions:
The novel pipeline introduced here that estimates functional and structural network disruptions from lesion masks demonstrates that it may provide a time-efficient yet behaviorally relevant alternative to advanced MRI in patient populations. This pipeline offers clinicians a practical tool for better understanding how lesions impact brain networks, and enable more accurate prognosis and recovery planning. This approach preserves inter-individual differences, enabling personalized predictions of motor recovery and enhancing the potential for tailored interventions in stroke rehabilitation.
Learning and Memory:
Neural Plasticity and Recovery of Function
Modeling and Analysis Methods:
Classification and Predictive Modeling
Connectivity (eg. functional, effective, structural) 2
fMRI Connectivity and Network Modeling
Motor Behavior:
Motor Behavior Other 1
Keywords:
Computational Neuroscience
Motor
Neurological
Other - stroke; Motor Recovery; Structural Connectome (SC); Functional Connectome (FC); Lesion; outcome prediction; neurovascular
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.
Jamison, K. W., Gu, Z., Wang, Q., Tozlu, C., Sabuncu, M. R., & Kuceyeski, A. (2024). Release the Krakencoder: A unified brain connectome translation and fusion tool. bioRxiv : the preprint server for biology, 2024.04.12.589274.
Kuceyeski, A., Maruta, J., Relkin, N., & Raj, A. (2013). The Network Modification (NeMo) Tool: elucidating the effect of white matter integrity changes on cortical and subcortical structural connectivity. Brain connectivity, 3(5), 451–463.