Investigation of functional MRI using phase-based EPT: Comparison with simulations
Poster No:
2618
Submission Type:
Abstract Submission
Authors:
Kyu-Jin Jung1, Chuanjiang Cui1, SooHyoung Lee1, Ji-Won Chun2, Dong-Hyun Kim1
Institutions:
1Department of Electrical and Electronic Engineering, Yonsei University, Seoul, Korea, Republic of, 2Department of Medical Informatics, Catholic University of Korea College of Medicine, Seoul, Korea, Republic of
First Author:
Kyu-Jin Jung
Department of Electrical and Electronic Engineering, Yonsei University
Seoul, Korea, Republic of
Department of Electrical and Electronic Engineering, Yonsei University
Seoul, Korea, Republic of
Co-Author(s):
Chuanjiang Cui
Department of Electrical and Electronic Engineering, Yonsei University
Seoul, Korea, Republic of
Department of Electrical and Electronic Engineering, Yonsei University
Seoul, Korea, Republic of
SooHyoung Lee
Department of Electrical and Electronic Engineering, Yonsei University
Seoul, Korea, Republic of
Department of Electrical and Electronic Engineering, Yonsei University
Seoul, Korea, Republic of
Ji-Won Chun
Department of Medical Informatics, Catholic University of Korea College of Medicine
Seoul, Korea, Republic of
Department of Medical Informatics, Catholic University of Korea College of Medicine
Seoul, Korea, Republic of
Dong-Hyun Kim
Department of Electrical and Electronic Engineering, Yonsei University
Seoul, Korea, Republic of
Department of Electrical and Electronic Engineering, Yonsei University
Seoul, Korea, Republic of
Introduction:
Functional magnetic resonance imaging (fMRI) provides insight into brain activity by the blood oxygenation level-dependent (BOLD) effect1. These changes are usually detectable through the gradient-recalled echo-planar imaging (EPI).
Beyond the endeavors to enhance BOLD signals, significant efforts for another aspect have been made to decipher the physiological mechanisms underpinning the BOLD effect process2-4. On the other hand, only a handful of studies have attempted to clarify the relationship between brain activation and conductivity by applying electrical properties tomography (EPT) algorithms4-6. Despite these few attempts, a clear correlation remains yet elusive, with some studies even showing conflicting results5,6.
Motivated by this ambiguity, we embarked on an investigation of the relationship between BOLD signals and conductivity changes, which can be reconstructed from the B1 phase with bSSFP sequence, employing the phase-based EPT algorithm7,8. Additionally, we analyzed trends in both phase and conductivity information using RF simulation, which provides ground-truth.
Beyond the endeavors to enhance BOLD signals, significant efforts for another aspect have been made to decipher the physiological mechanisms underpinning the BOLD effect process2-4. On the other hand, only a handful of studies have attempted to clarify the relationship between brain activation and conductivity by applying electrical properties tomography (EPT) algorithms4-6. Despite these few attempts, a clear correlation remains yet elusive, with some studies even showing conflicting results5,6.
Motivated by this ambiguity, we embarked on an investigation of the relationship between BOLD signals and conductivity changes, which can be reconstructed from the B1 phase with bSSFP sequence, employing the phase-based EPT algorithm7,8. Additionally, we analyzed trends in both phase and conductivity information using RF simulation, which provides ground-truth.
Methods:
Six volunteers were scanned on a 3T MRI system (MAGNETOM Vida, Siemens Healthineers). For each volunteer, 2D EPI and bSFFP sequences were used (described in Figure 1 (A)). The phase information (φ) from bSSFP was used to reconstruct conductivity maps by employing the phase-based EPT algorithm8.
In the right finger-tapping experiment, we used a block design for both sequences, consisting of 80 dynamics. These were divided into four alternating blocks of resting and tasking states. Initially, EPI scan was performed on the subject to map the BOLD signal, following which we selected a specific slice where activated BOLD signals were noticeable. This identical slice was subsequently scanned using the bSSFP sequence with signal averaging twelve-times to ensure sufficient SNR levels for the EPT reconstruction.
Using GLM, we estimated beta maps for BOLD characteristics at p<0.05. To address uncertainties in B1 phase and conductivity activation function, we employed correlation analyses with a significance of r<0.05.
Additionally, we used the finite-difference time-domain electromagnetic simulation program, Sim4Life, to investigate the relationship between the phase and conductivity due to brain activation in ideal conditions.
In the right finger-tapping experiment, we used a block design for both sequences, consisting of 80 dynamics. These were divided into four alternating blocks of resting and tasking states. Initially, EPI scan was performed on the subject to map the BOLD signal, following which we selected a specific slice where activated BOLD signals were noticeable. This identical slice was subsequently scanned using the bSSFP sequence with signal averaging twelve-times to ensure sufficient SNR levels for the EPT reconstruction.
Using GLM, we estimated beta maps for BOLD characteristics at p<0.05. To address uncertainties in B1 phase and conductivity activation function, we employed correlation analyses with a significance of r<0.05.
Additionally, we used the finite-difference time-domain electromagnetic simulation program, Sim4Life, to investigate the relationship between the phase and conductivity due to brain activation in ideal conditions.
Results:
In Figure 1 (B), using EPI BOLD as the reference, bSSFP BOLD, phase, and conductivity changes were observed for in-vivo. Phase change showed a positive correlation around the motor cortex, similar to the trends observed in BOLD, while reconstructed conductivity demonstrated a negative correlation. Similar to the correlation results observed in in-vivo studies, comparable trends were also noted in the simulation, as depicted in Figure 1 (C).
Figure 2 presents the averaged results of the temporal-series signals in the non-activated (right motor cortex) and activated (left motor cortex) across six subjects. BOLDs and phase showed an increase in temporal series during the activation state, while conductivity exhibited a decrease. These trends of change are more pronounced when comparing activated and non-activated regions, with phase showing an increase of 0.017% in the signal during activation, while conductivity decreased by 0.37%.
Figure 2 presents the averaged results of the temporal-series signals in the non-activated (right motor cortex) and activated (left motor cortex) across six subjects. BOLDs and phase showed an increase in temporal series during the activation state, while conductivity exhibited a decrease. These trends of change are more pronounced when comparing activated and non-activated regions, with phase showing an increase of 0.017% in the signal during activation, while conductivity decreased by 0.37%.
Conclusions:
We explored the phase and conductivity changes during brain activation. Similar activation trends in in-vivo and simulation were observed in both phase and conductivity. This study considered how brain activation might affect conductivity, potentially influenced by ion concentration and blood oxygenation. It is known that ion concentration increases conductivity7, but other factors (such as red cell9/hemoglobin concentrations10) may impact this effect. This conflict during brain activation highlights the need to understand which factor more significantly impacts conductivity changes.
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI)
Methods Development
Other Methods
Novel Imaging Acquisition Methods:
Imaging Methods Other 2
Physiology, Metabolism and Neurotransmission :
Neurophysiology of Imaging Signals 1
Keywords:
Computational Neuroscience
Data analysis
Experimental Design
fMRI CONTRAST MECHANISMS
FUNCTIONAL MRI
MRI PHYSICS
NORMAL HUMAN
Other - Electrical Properties Tomography
1|2Indicates the priority used for review
Provide references using author date format
[2] Ip, I.B., et al. (2017), ‘Combined fMRI-MRS acquires simultaneous glutamate and BOLD-fMRI signals in the human brain’, Neuroimage, Vol. 155, no. 15, pp. 113-119.
[3] Balla, D.Z., et al. (2014), ‘Functional quantitative susceptibility mapping (fQSM)’, Neuroimage, Vol 100, no. 15, pp. 112-124.
[4] Helle M. & Katscher U (2019), ‘Electrical Properties Tomography based Functional Magnetic Resonance Imaging (EPT-fMRI)’, In proceedings of ISMRM 27th Annual Meeting & Exhibition, p. 3759.
[5] Jung, K.J., et al. (2023), ‘Investigation of electrical conductivity changes during functional activity of the brain via phase-based MR-EPT: Preliminary observation’, In proceedings of 2023 ISMRM Annual Meeting & Exhibition, p. 0922.
[6] Jahng, G.H., et al. (2023), ‘High-frequency Conductivity Signal Changes Measured with Functional MREPT during Visual Stimulation’, In proceedings of 2023 ISMRM Annual Meeting & Exhibition, p. 0923.
[7] Katscher U. & van den Berg C.A.T. (2017), ‘Electric properties tomography: Biochemical, physical and technical background, evaluation and clinical applications’, NMR Biomed. Vol. 30, no. 8, p. 3729.
[8] Lee, J., Shin, J., & Kim, D.H. (2016), ‘MR‐based conductivity imaging using multiple receiver coils’, Magnetic resonance in medicine, Vol. 76, no. 2, pp. 530-539.
[9] Hirsch F.G., et al. (1950) , ’The electrical conductivity of blood: 1. Relationship to erythrocyte concentration’, Blood, Vol. 5, no. 11, pp. 1017-1035.
[10] Altaf I.M. & Ahmad A. (2017), ‘Dielectric properties of human blood of patients Suffering from iron deficiency anemia’, International Journal of Science, Environment and Technology, Vol. 6, no. 1, pp. 916-922.