3134

Enhancing Magnetization Transfer Contrast (MTC): A proof-of-concept study of MTC-STAGE Imaging

Jun Chen Li^1,2, Yu Liu¹, Yongsheng Chen³, Zhijia Jin¹, Naying He¹, Weibo Chen⁴, Fuhua Yan¹, and Ewart Mark Haacke^1,5,6
¹Radiology, Ruijin Hospital, Shanghai, China, ²Radiology, Changshu Hospital Affiliated to Nanjing University of Chinese Medicine, Changshu, China, ³Neurology, Wayne State University, Detroit, MI, United States, ⁴Philips Healthcare, Shanghai, China, ⁵Radiology, Wayne State University, Detroit, MI, United States, ⁶Biomedical Engineering, Wayne State University, Detroit, MI, United States

Synopsis

Magnetization transfer contrast (MTC) imaging has been used to study neuromelanin (NM) in Parkinson’s disease. By suppressing the background tissue using an MTC pulse in a T1W sequence, the NM becomes visible, supposedly because of its reduced T1. However, we show using STAGE (strategically acquired gradient echo) imaging with/without an MTC pulse that this is not the reason for its visibility. Rather, it is the increased water content relative to surrounding tissue that keeps the signal high. Using the appropriate choice of flip angles and resolution, the NM contrast on MTC images can be significantly increased.

Introduction:

Parkinson’s disease (PD) is one of the fastest growing neurological disorders¹. Diagnosing PD from other movement disorders such as essential tremor is an important part of determining patient treatment². Recently, there has been a major effort to use neuromelanin (NM) imaging as a biomarker for diagnosing PD. It is well known that there is a loss of NM in patients with PD and that upon presentation with symptoms the patients have already lost 50% to 60% of their NM³. Magnetization transfer contrast (MTC) imaging has been used to image NM for more than a decade now by suppressing the background tissue in T1W sequences. The presence of macromolecules in the white matter and to a lesser degree in gray matter causes both their signals to be significantly decreased leaving tissue with more water having the highest signal. Most papers published to date suggest that the NM becomes visible because of its reduced T1. However, we show using strategically acquired gradient echo (STAGE)^4-6 imaging that this is not the major reason for its visibility. Rather, it is the increased water content relative to the surrounding tissue that keeps the signal high for the NM in the substantia nigra (SN). Using the known tissue properties makes it possible to use a spin density weighted MTC acquisition to further enhance the NM contrast. Therefore, we propose using a MTC-STAGE susceptibility weighted imaging (SWI) like data acquisition with both proton spin density weighting (PSDW) and T1 weighting (T1W) to optimize the NM contrast. The purpose of this study was to quantify the tissue properties of the midbrain and NM using a two flip angle STAGE approach in an MTC-SWI sequence to measure the tissue properties both with and without the MTC radiofrequency pulse. Finally, we also create a new image using the product of these two sequences (pMTC) to enhance NM visibility.

Methods:

A total of 4 healthy controls (ages 30, 20, 20, 23 years old) were scanned on a 3T Philips scanner using a 3D multi-echo gradient echo SWI sequence with magnetization transfer contrast (MTC). The imaging parameters included: seven echoes with TE1 = 7.5ms, ΔTE = 7.5ms, with TR = 62ms, pixel bandwidth = 174Hz/pixel, matrix size = 384 × 384, slice thickness = 2mm, and a spatial in-plane resolution = 0.67 × 0.67mm². The first volunteer (30 years old) was scanned with flip angles (FA) of 15 ˚ and 40˚ and STAGE processing was performed on this data to predict the results from a range of FAs. In the last 3 volunteers, a series of flip angles was run ranging from 5°, 8°, 10°, 12°, 15°, 20° and 30°. The first echo in the MTC-SWI magnitude image (TE = 7.5ms) was used to delineate the NM content in the midbrain. The MTC pulse parameters were: FA=90°, offset zero, and 3 block pulses of 1.914 ms. Contrast measurements were made in three regions: in the NM visible area itself, above the SN in the cerebral peduncle and below the SN in the adjacent gray matter.

Results:

Figure 1 shows the original MTC results with 15° and 40° FAs, and STAGE processing results including: true-PSD (or tPSD) map, and QSM. The T1 map and PSD maps demonstrated that the MT contrast mainly occurs because of the high proton spin density in the NM. Overall contrast for the NM is higher in the 15° flip angle image but the cerebrospinal fluid appears very bright. The 15° FA image shows the cerebral peduncle darker than the 40° FA image. This gives the appearance of better contrast in the anterior part of the NM territory while the posterior part has better contrast for the 40° FA image. The product of these two images (pMTC in Figure 1), gives a compromise exhibiting good contrast in both regions. The Ernst equation plot for the signal intensity as a function of flip angle nicely matches the data’s contrast (Figure 2). Moreover, the maximum MT contrast is reached at 12°, but 15° which agrees well with the data (Figures 3 and 4). Using 15° and 40° for MTC-STAGE. gives more than 80% of the maximum T1 precision (usually reached at [10°, 55°] and [9°, 48°], for WM and GM).

Discussion and Conclusion:

In this work, we have introduced a new approach to enhance NM contrast using a two flip angle MTC-STAGE protocol to understand the tissue properties. We have shown that both 15° and 20° can provide good contrast in the upper and lower region of the NM. A combination of both a low and high flip image can be used to enhance NM contrast in the product image pMTC. The lower T1 is actually detrimental to the contrast at higher flip angles because contrast there diminishes due to the higher signal of high water content longer T1 components in the NM. Therefore, the main reason for the enhanced NM contrast comes from the increased water content in the NM and low FAs provide the best contrast. This has potential applications for all studies related to using NM as a biomarker to better diagnose Parkinson’s patients.

Acknowledgements

No acknowledgement found.

References

1. Group GBDNDC. Global, regional, and national burden of neurological disorders during 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet Neurol 2017; 16(11): 877-97.

2. Damier P et al. The substantia nigra of the human brain. II. Patterns of loss of dopamine-containing neurons in Parkinson's disease. Brain 1999; 122 ( Pt 8): 1437-48.

3. Zheng ZH et al. Imaging the Nigrosome 1 in the substantia nigra using SWI and QSM: An application to Parkinson’s disease. Neuroimage Clinical: in press.

4. Chen Y, Liu S, Wang Y, Kang Y, Haacke EM. STrategically Acquired Gradient Echo (STAGE) imaging, part I: Creating enhanced T1 contrast and standardized susceptibility weighted imaging and quantitative susceptibility mapping. Magnetic resonance imaging 2018;46:130-139.

5. Wang Y, Chen Y, Wu D, et al. STrategically Acquired Gradient Echo (STAGE) imaging, part II: Correcting for RF inhomogeneities in estimating T1 and proton density. Magnetic resonance imaging 2018;46:140-150.

6. Haacke EM, Chen Y, Utrainen D, et al. STrategically Acquired Gradient Echo (STAGE) Imaging, part III: Technical Advances and Clinical Applications of A Rapid Multi-Contrast Multi-Parametric Brain Imaging Method. Magnetic Resoannce Imaging 2019;DOI:10.1016/j.mri.2019.09.006.

Figures

Figure 1. Representative images from MTC-STAGE for a 30 year old healthy male. Two scans were acquired with MTC: FA=15° and FA=40°. The product of these to MTC images (pMTC) gives a compromise exhibiting good contrast in both the upper and lower regions. Meanwhile, the STAGE processing gave SWI, tSWI, QSM and a T1WE image (neither of the two FA MTC datasets gave T1W images) as well as four quantitative data sets: a T1 map, T2* weighted PSD map, true-PSD (or tPSD) map, and QSM. The T1 map and PSD maps demonstrated that the MT contrast mainly occurs because of the high proton spin density in the NM.

Figure 2. MTC-STAGE data (A-E). STAGE T1/PSD processing was performed using 15° (A) and 40° (B) images. The measured T1 value of the genu of the corpus callosum (CC) was 670 ms, and of the right putamen was 910 ms (C). The PSD ratio was 0.8 (D). The STAGE T1WE (E) image is able to reveal a conventional T1W like contrast (not possible with conventional MTC imaging). The signal intensity as a function of flip angle nicely matches the data (F). Maximum MT contrast is reached around 12° to 15° and but 15° has much higher SNR. Therefore, we chose 15° and 40° for STAGE imaging to extract T1 and PSD.

Figure 3. CNR for the acquired and synthetic MTC (sMTC) images. The T1/PSD maps were generated from the MTC 15° (B) and 40° (C) images. The sMTC images were generated using the Ernst equation. Six ROIs (A) were manually drawn on left/right NM (rNM, lNM), and homogenous regions on the top/bottom of the NM regions for both sides (ltROI, rtROI, lbROI, rbROI) on the 20th slice. CNR was measured between the NM and the adjacent tissues (right side rNM and right-top rtROI). On the plot (G), the real data are shown in black and pMTC (spMTC) data at the origin (45°) both are in agreement with the predictions.

Figure 4. The contrast for the flip angles 5°, 8°, 10°, 12°, 15°, 20° and 30° is plotted for the same slice shown in Figures 1-3 for the 3 volunteers. The 3 left and right contrasts were averaged to find the mean over these 6 points. The error bars represent the standard error of the mean. The upper region is seen to peak at 15° and then rapid drops after this while the lower region peaks at 20°. Both of these observations are in good agreement with the simulated data.

Proc. Intl. Soc. Mag. Reson. Med. 28 (2020)

3134