Ultrafast functional magnetic resonance imaging with MR-encephalography
Nuklearmedizin, Strahlentherapie, Strahlenbiologie
Zusammenfassung der Projektergebnisse
The project has been aimed at the development of methods and technologies for ultrafast functional imaging of the brain based on the MR-Encephalography (MREG) sequence developed by the applicant, which allows to measure a full brain dataset of matrix 64x64x64 in 100 ms. The project has three main workpackages, namely hardware and technologies, methods, and applications: Technological developments are aimed at the development of a mulitcoil-array for signal reception consisting of 95 individual coil elements covering the brain. Compared to the existing coil this leads to a more than twofold improvement of the number of coil elements with the promise to improve the spatial localization especially in cortical regions close to the surface of the brain. Methodological developments include on one hand developments of the measurement sequence – a single-shot-stack of spiral-(ssSos)-sequence. An interleaved version of the sequence has been developed which allows a flexible compromise between image quality and measurement time. Furthermore a spin-echo variant of the sequence has been realized, which also yields better image quality and allows to include T2-contrast into the signal formation mechanism. Finally, k-space trajectories leading to two different echotimes have been realized in order to investigate the underlying physiological contributions to the observed correlated activities in resting state experiments. The standard implementation of MREG requires reconstruction of the highly undersampled data using an iterative regularization with L1/L2-norm and off-resonance correction. Full reconstruction requires long reconstruction times and has to be performed off-line. By developing a targeted reconstruction approach a real-time implementation of the method could be implemented, which allows to reconstruct on the fly signal from 100 predefined and arbitrarily shaped subvolumes within the 100 ms repetition time of the sequence, which allows the use of the method for realtime feedback applications. Speedup of reconstruction of the full dataset has been achieved using the Time-domain principal component reconstruction (tPCR), which is a very generally applicable method by which first a Principle Component Analysis is performed on the time domain data. Iterative reconstruction of the principal components has been shown to be much faster than framewise reconstruction and allows to improve data quality by including only the most relevant components in the final result The high sampling rate of MREG has been shown to allow separation of ECG- und breathing-related signal variations and thus to quite dramatically increase the sensitivity to detect signal changes related to neuronal activity. In the investigation of resting state networks this translates into faster detection of networks with more subnetworks to be detected within a given measurement time. The MREG-sequence has been applied in quite a number of studies both within our group as well as with clinical partners at our institution (neurosurgery (J. Jacobs), psychiatry (LT van Elst), neurology (C. Weiller), gender studies (A. Kaiser)) as well as with external partners. A very surprising and un-anticipated development was the work by V. Kiviniemi on the observed cardioballistic pulsatility, which is considered a nuisance for fMRI studies. He could show very significant changes in the spatiotemporal patterns of ECG - and breathing related pulsatility in different patient groups – including Alzheimers patients. The observations are in line with the hypothesized role of the glymphatic system which has been postulated by Maiken Nedergaard. The MREG-sequence has been distributed to 20+ sites around the world. Among others fruitful collaborations have been built up with the University of Maastricht (R. Goebel) for realtime fMRI and with the University of Oulu, Finnland (V. Kiviniemi), which has performed multiple studies using MREG for the assessment and characterization of the glymphatic system, which is hypothesized to be involved in the clearing of the brain from metabolic byproducts and has been shown to play a role in various disease including Alzheimers disease.
Projektbezogene Publikationen (Auswahl)
- EEG-fMRI Gradient Artifact Correction by Multiple Motion-Related Templates. IEEE Trans Biomed Eng. 63, 2647–2653 (2016)
P. LeVan, S. Zhang, B. Knowles, M. Zaitsev, J. Hennig
(Siehe online unter https://doi.org/10.1109/tbme.2016.2593726) - Marker-based ballistocardiographic artifact correction improves spike identification in EEG-fMRI of focal epilepsy patients. Clin Neurophysiol. 127, 2802–2811 (2016)
K. Körbl, J. Jacobs, M. Herbst, M. Zaitsev, A. Schulze-Bonhage, J. Hennig, P. LeVan
(Siehe online unter https://doi.org/10.1016/j.clinph.2016.05.361) - Enhanced subject-specific resting-state network detection and extraction with fast fMRI. Hum Brain Mapp. 38, 817–830 (2017)
B. Akin, H. L. Lee, J. Hennig, P. LeVan
(Siehe online unter https://doi.org/10.1002/hbm.23420) - Targeted partial reconstruction for real-time fMRI with arbitrary trajectories. Magn. Reson. Med. 81, 1118–1129 (2019)
B.Riemenschneider, P. Levan, J. Hennig
(Siehe online unter https://doi.org/10.1002/mrm.27478) - The variability of functional MRI brain signal increases in Alzheimer’s disease at cardiorespiratory frequencies. Sci Rep 2020;10:21559
Tuovinen T, Kananen J, Rajna Z, et al.
(Siehe online unter https://doi.org/10.1038/s41598-020-77984-1) - Time-domain principal component reconstruction (tPCR): A more efficient and stable iterative reconstruction framework for non-Cartesian functional MRI. Magn. Reson. Med. 2020;84:1321–1335 d
Wang F, Hennig J, LeVan P
(Siehe online unter https://doi.org/10.1002/mrm.28208) - 15 Years MR-encephalography. MAGMA 2021;34:85–108
Hennig J, Kiviniemi V, Riemenschneider B, et al.
(Siehe online unter https://doi.org/10.1007/s10334-020-00891-z) - Improving the sensitivity of spinecho fMRI at 3T by highly accelerated acquisitions. Magn Reson Med 2021;86:245–257
Barghoorn A, Riemenschneider B, Hennig J, LeVan P
(Siehe online unter https://doi.org/10.1002/mrm.28715) - Trading off spatio-temporal properties in 3D high-speed fMRI using interleaved stack-of-spirals trajectories. Magn Reson Med 2021;86:777–790
Riemenschneider B, Akin B, LeVan P, Hennig J
(Siehe online unter https://doi.org/10.1002/mrm.28742) - Multi-Echo MR-Encephalography using Spherical Stack of Spirals Trajectories. Proc. XXI Annual Meeting ISMRM, p. 1102 (2022)
Barghoorn, A, Riemenschneider, B, Yang, Wenchao, Hennig, J
