Evaluation of a Fast Real-Time Echo-Combination Method for Multi-Echo fMRI
Poster No:
1879
Submission Type:
Abstract Submission
Authors:
Brice Fernandez1, Alexander Cohen2, Philipp Sämann3, Michael Czisch3, Yang Wang2, Victor Spoormaker3
Institutions:
1GE HealthCare, Buc, France, 2Medical College of Wisconsin, Milwaukee, WI, 3Max Planck Institute of Psychiatry, Munich, Germany
First Author:
Co-Author(s):
Introduction:
To take advantage of ME-fMRI for real-time (RT) applications, it requires to generate an Echo Combined (EC) image in RT (Heunis, 2021) usually as a normalized weighted sum. Several methods were proposed to estimate these weights including: the T2*-weighted Optimal Echo Combination (OptEC) (Posse, 1999), and a per TR T2*FIT (Heunis, 2021) which can be seen as the RT version of OptEC.
The T2*FIT combination would be the natural choice for RT applications, but it requires a T2* fitting (per voxel per TR in RT) which makes it computationally intensive.
Here, a fast RT Echo Combination method (named RTEC) that takes advantage of T2* decay information naturally contained in ME data is suggested and evaluated against OptEC on two datasets with task and rest.
The T2*FIT combination would be the natural choice for RT applications, but it requires a T2* fitting (per voxel per TR in RT) which makes it computationally intensive.
Here, a fast RT Echo Combination method (named RTEC) that takes advantage of T2* decay information naturally contained in ME data is suggested and evaluated against OptEC on two datasets with task and rest.
Methods:
The suggested RTEC is described by the equation on fig 1-A. We have Ne echoes and the nth echo En is described by eq 1, EC is given by eq 2 and the weights satisfy eq 3. The optimal weights (wnopt) are given by eq 4 where T2* is an estimate of T2* (usually based on the mean timeseries). Then, if the numerator and denominator of wnopt are multiplied by E0 (signal at TE=0), then the En can be identified, and the formula simplified as eq 5. Hence, the weights can be computed based only on En and TEn which is convenient for RT applications. To be best of our knowledge, this simpler formulation has not yet been suggested and evaluated.
The performance of RTEC against OptEC was evaluated offline based on existing 2 datasets. The 1st was composed of a fear conditioning task and resting state fMRI on 32 subjects (Fernandez, 2017) - 3 echoes, TE=12/29/46 ms, TR=2.56 s. The 2nd included a visual checkerboard and resting state fMRI on 28 subjects (Cohen, 2021a/b) - 3 echoes TE=11/30/49 ms, TR=0.9 s, multiband=4.
The fear-conditioning task is composed of conditioned stimuli (CS+) and a safe stimulus (CS-). The comparison CS->CS+ shows the ventromedial prefrontal cortex (vmPFC) while CS+>CS- shows the dorsal anterior cingulate (dACC) and bilateral insula among others (Fullana, 2015).
The analysis pipeline was designed to compare OptEC vs RTEC without introducing undesired variability. Family wise error correction at cluster level (pc) or voxel level (pv) was used for multiple testing correction.
The performance of RTEC against OptEC was evaluated offline based on existing 2 datasets. The 1st was composed of a fear conditioning task and resting state fMRI on 32 subjects (Fernandez, 2017) - 3 echoes, TE=12/29/46 ms, TR=2.56 s. The 2nd included a visual checkerboard and resting state fMRI on 28 subjects (Cohen, 2021a/b) - 3 echoes TE=11/30/49 ms, TR=0.9 s, multiband=4.
The fear-conditioning task is composed of conditioned stimuli (CS+) and a safe stimulus (CS-). The comparison CS->CS+ shows the ventromedial prefrontal cortex (vmPFC) while CS+>CS- shows the dorsal anterior cingulate (dACC) and bilateral insula among others (Fullana, 2015).
The analysis pipeline was designed to compare OptEC vs RTEC without introducing undesired variability. Family wise error correction at cluster level (pc) or voxel level (pv) was used for multiple testing correction.
·Figure 1
Results:
The relative difference (mean±std%) between the EC images (fig 1-B) shows that the difference is small (<1±1%) for most brain regions except for the susceptibility affected regions (e.g. vmPFC) where it reaches 5±4%. All expected relevant cluster of the fear conditioning task were detected with both OptEC and RTEC (fig 1-C). However, the cluster in the vmPFC with RTEC was much smaller (1.2 vs 7.8 cm3). A paired t-test revealed that all clusters were significantly stronger at pc<0.05 for OptEC. The expected cluster of the visual checkerboard task were detected with both method (fig 1-D). A paired t-test revealed that the cluster shown in fig. 1-D was significantly stronger for OptEC.
On the resting-state of dataset 1, a vmPFC seed analysis was performed for OptEC and RTEC (fig 2-A). As expected, OptEC showed all the DMN nodes (pv<0.05) while RTEC did not until the threshold is markedly lowered (pc<0.05). A paired t-test OptEC>RTEC revealed a cluster in the vmPFC at pc<0.05. A PCC seed-based analysis (dataset 2) revealed the DMN but one cluster on the right amygdala was missing for OptEC. A paired t-test OptEC>RTEC revealed a cluster in the vmPFC (pc=0.043).
On the resting-state of dataset 1, a vmPFC seed analysis was performed for OptEC and RTEC (fig 2-A). As expected, OptEC showed all the DMN nodes (pv<0.05) while RTEC did not until the threshold is markedly lowered (pc<0.05). A paired t-test OptEC>RTEC revealed a cluster in the vmPFC at pc<0.05. A PCC seed-based analysis (dataset 2) revealed the DMN but one cluster on the right amygdala was missing for OptEC. A paired t-test OptEC>RTEC revealed a cluster in the vmPFC (pc=0.043).
·Figure 2
Conclusions:
These results indicate that the RTEC approach is a viable choice if computational power is limited. RTEC was successful in detecting reliably activation and connection in most part of the brain except in the vmPFC where its performance seems more severely impacted.
Future work will include comparison with other RT method such as T2*FIT (Heunis, 2021).
Future work will include comparison with other RT method such as T2*FIT (Heunis, 2021).
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI)
Connectivity (eg. functional, effective, structural)
Methods Development 2
Novel Imaging Acquisition Methods:
BOLD fMRI 1
Keywords:
Acquisition
fMRI CONTRAST MECHANISMS
FUNCTIONAL MRI
MRI
MRI PHYSICS
Other - Multi-echo fMRI
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.
Cohen, A. D., Yang, B., Fernandez, B., Banerjee, S., & Wang, Y. (2021b). Improved resting state functional connectivity sensitivity and reproducibility using a multiband multi-echo acquisition. NeuroImage, 225, 117461. https://doi.org/10.1016/j.neuroimage.2020.117461
Fernandez, B., Leuchs, L., Sämann, P. G., Czisch, M., & Spoormaker, V. I. (2017). Multi-echo EPI of human fear conditioning reveals improved BOLD detection in ventromedial prefrontal cortex. NeuroImage, 156, 65–77. https://doi.org/10.1016/j.neuroimage.2017.05.005
Fullana, M. A., Harrison, B. J., Soriano-Mas, C., Vervliet, B., Cardoner, N., Àvila-Parcet, A., & Radua, J. (2016). Neural signatures of human fear conditioning: an updated and extended meta-analysis of fMRI studies. Molecular psychiatry, 21(4), 500–508. https://doi.org/10.1038/mp.2015.88
Heunis, S., Breeuwer, M., Caballero-Gaudes, C., Hellrung, L., Huijbers, W., Jansen, J. F., Lamerichs, R., Zinger, S., & Aldenkamp, A. P. (2021). The effects of multi-echo fMRI combination and rapid T2*-mapping on offline and real-time BOLD sensitivity. NeuroImage, 238, 118244. https://doi.org/10.1016/j.neuroimage.2021.118244
Poser, B. A., Versluis, M. J., Hoogduin, J. M., & Norris, D. G. (2006). BOLD contrast sensitivity enhancement and artifact reduction with multiecho EPI: parallel-acquired inhomogeneity-desensitized fMRI. Magnetic resonance in medicine, 55(6), 1227–1235. https://doi.org/10.1002/mrm.20900
Posse, S., Wiese, S., Gembris, D., Mathiak, K., Kessler, C., Grosse-Ruyken, M. L., Elghahwagi, B., Richards, T., Dager, S. R., & Kiselev, V. G. (1999). Enhancement of BOLD-contrast sensitivity by single-shot multi-echo functional MR imaging. Magnetic resonance in medicine, 42(1), 87–97. https://doi.org/10.1002/(sici)1522-2594(199907)42:1<87::aid-mrm13>3.0.co;2-o