Making high-resolution brain MRI images easier to read with the FLAWS sequence #ST55 [fr]
A cooperation between the Signal and Image Processing Laboratory (LTSI) in Rennes and the Australian e-Health Research Centre (AEHRC) of the CSIRO in Brisbane has implemented a new method to facilitate the reading of MRI images of the brain. The study, recently published in the scientific journal Magnetic Resonance in Medicine, benefited from the French expertise in MRI physics of the LTSI, and from the Australian expertise in brain imaging of the AEHRC.
Magnetic Resonance imaging (MRI) is a non-invasive imaging technique used to visualise the human body by detecting responses of tissues to magnetic fields. MRI provides very good contrast between soft tissues, compared to structural imaging methods obtained from X-rays, such as computed tomography. For this reason, MRI is the reference method for brain imaging.
The MRI sequence Fluid and white matter suppression (FLAWS) provides several anatomical images of the brain: an image with white matter suppression, an image with cerebrospinal fluid suppression, and an image specific to grey matter. These different images are used to better visualise brain damage, especially in the cortex, for pathologies such as epilepsy. FLAWS imaging also provides good visualisation of the central grey nuclei, structures located deep in the cerebral hemispheres and which are involved in learning and controlling motor movements. This technique could therefore improve the planning of surgery operations for deep brain stimulation to treat diseases such as Parkinson’s disease.
Very high magnetic field MRI, for which the magnetic field rises to 7 Tesla (7T imaging), provides high resolution images (less than 1mm3) that provide more details on the different brain tissues, compared to high magnetic field imaging such as 3 Tesla imaging, which is used in clinical routine. However, 7T imaging is strongly impacted by magnetic field inhomogeneity effects, making it more difficult for radiologists to read the images and complicating the measurement of magnetic properties of brain tissue, such as the measurement of T1 relaxation times which vary according to tissues composition.
This French-Australian collaborative study results from the work of Dr. Jérémy Beaumont, who carried out his thesis in co-direction between the LTSI and the AEHRC. Titled “High-resolution multi-T1-weighted contrast and T1 mapping with low B1+ sensitivity using the fluid and white matter suppression sequence at 7T”, the study proposes an optimisation of the MRI FLAWS sequence at 7T to obtain anatomical images of the brain that are less sensitive to the effects of magnetic field inhomogeneities. In this study, a new method of measuring T1 relaxation times is introduced to obtain results that are less sensitive to magnetic field inhomogeneities, while ensuring high-resolution imaging. These new advances in ultra-high magnetic field imaging could improve the detection of brain damage for diseases such as epilepsy, and could also increase the reproducibility of T1 relaxation time measurements of brain tissue at high-resolution to better characterise brain pathologies.
Figure 1. Examples of FLAWS images obtained at 7T. The FLAWS-hc and FLAWS-hco images are reconstructed from the acquired FLAWS1 and FLAWS2 images to obtain images with low sensitivity to magnetic field inhomogeneities. This reduced sensitivity to magnetic field inhomogeneities is particularly visible when comparing the white matter visualisation between FLAWS2 and FLAWS-hco (brain tissues appearing in white in these images): the white substance signal changes intensity (color change) depending on its location in FLAWS2, while this signal remains constant (no color change) depending on its location in FLAWS-hco. This reduced sensitivity to magnetic field inhomogeneities provided by FLAWS-hc and FLAWS-hco facilitates image interpretation for radiologists. Images courtesy of Dr. J. Beaumont.
Figure 2. Examples of T1 relaxation time mappings obtained with a 7T reference sequence (called MP2RAGE method, top figures) and a FLAWS sequence at 7T (bottom figures). The MP2RAGE sequence is a reference sequence in terms of T1 relaxation time mapping with low sensitivity to magnetic field inhomogeneities at 7T. In this figure, we find that the FLAWS sequence provides a mapping of T1 relaxation times with low sensitivity to magnetic field inhomogeneities, which is characterised by a better resolution (0.5 mm3) than the mapping provided by the MP2RAGE sequence (1.0 mm3). Images courtesy of Dr. J. Beaumont.