Standardized Data Acquisition for Neuromelanin-Sensitive Magnetic Resonance Imaging of the Substantia Nigra

Summary

This protocol shows how to acquire neuromelanin-sensitive magnetic resonance imaging data of the substantia nigra.

Abstract

The dopaminergic system plays a crucial role in healthy cognition (e.g., reward learning and uncertainty) and neuropsychiatric disorders (e.g., Parkinson's disease and schizophrenia). Neuromelanin is a byproduct of dopamine synthesis that accumulates in dopaminergic neurons of the substantia nigra. Neuromelanin-sensitive magnetic resonance imaging (NM-MRI) is a noninvasive method for measuring neuromelanin in those dopaminergic neurons, providing a direct measure of dopaminergic cell loss in the substantia nigra and a proxy measure of dopamine function. Although NM-MRI has been shown to be useful for studying various neuropsychiatric disorders, it is challenged by a limited field-of-view in the inferior-superior direction resulting in the potential loss of data from the accidental exclusion of part of the substantia nigra. In addition, the field is lacking a standardized protocol for the acquisition of NM-MRI data, a critical step in facilitating large-scale multisite studies and translation into the clinic. This protocol describes a step-by-step NM-MRI volume placement procedure and online quality control checks to ensure the acquisition of good-quality data covering the entire substantia nigra.

Introduction

Neuromelanin (NM) is a dark pigment found in dopaminergic neurons of the substantia nigra (SN) and noradrenergic neurons of the locus coeruleus (LC)^1,2. NM is synthesized by the iron-dependent oxidation of cytosolic dopamine and norepinephrine and is stored in autophagic vacuoles in the soma³. It first appears in humans around 2-3 years of age and accumulates with age^1,4,5.

Within the NM-containing vacuoles of SN and LC neurons, NM forms complexes with iron. These NM-iron complexes are paramagnetic, allowing for noninvasive visualization of NM using magnetic resonance imaging (MRI)^6,7. MRI scans that can visualize NM are known as NM-sensitive MRI (NM-MRI) and use either direct or indirect magnetization transfer effects to provide contrast between regions with high NM concentration (e.g., the SN) and the surrounding white matter^8,9.

Magnetization transfer contrast is the result of the interaction between macromolecular-bound water protons (which are saturated by the magnetization transfer pulses) and the surrounding free water protons. In NM-MRI, it is believed that the paramagnetic nature of NM-iron complexes shortens the T₁ of the surrounding free water protons, resulting in reduced magnetization-transfer effects so that regions with higher NM concentration appear hyperintense on NM-MRI scans¹⁰. Conversely, the white matter surrounding the SN has a high macromolecular content, resulting in large magnetization-transfer effects so that these regions appear hypointense on NM-MRI scans, thus providing high contrast between the SN and surrounding white matter.

In the SN, NM-MRI can provide a marker of dopaminergic cell loss¹¹ and dopamine system function¹². These two processes are relevant for several neuropsychiatric disorders and are supported by a vast body of clinical and preclinical work. For example, abnormalities in dopamine function have been widely observed in schizophrenia; in vivo studies using positron emission tomography (PET) have shown increased striatal dopamine release^13,14,15,16 and increased dopamine synthesis capacity^{17,18,19,20,21,22}. Furthermore, post-mortem studies have shown that patients with schizophrenia have increased levels of tyrosine hydroxylase—the rate-limiting enzyme involved in dopamine synthesis—in the basal ganglia²³ and SN^24,25.

Several studies have investigated patterns of dopaminergic cell loss, particularly in Parkinson's disease. Post-mortem studies have revealed that the pigmented dopaminergic neurons of the SN are the primary site of neurodegeneration in Parkinson's disease^26,27, and that, while SN cell loss in Parkinson's disease is not correlated with cell loss in normal aging²⁸, it is correlated with the duration of the disease²⁹. Unlike most methods for investigating the dopaminergic system, the non-invasiveness, cost-effectiveness, and lack of ionizing radiation make NM-MRI a versatile biomarker³⁰.

The NM-MRI protocol described in this paper was developed to increase both within-subject and across-subject reproducibility of NM-MRI. This protocol ensures full coverage of the SN despite the limited coverage of NM-MRI scans in the inferior-superior direction. The protocol makes use of sagittal, coronal, and axial three-dimensional (3D) T1-weighted (T1w) images, and the steps should be followed to achieve proper slice stack placement. The protocol outlined in this paper has been utilized in multiple studies^31,32 and was extensively tested. Wengler et al. completed a study of the reliability of this protocol in which NM-MRI images were acquired twice in each participant across multiple days³². Intra-class correlation coefficients demonstrated excellent test-retest reliability of this method for region of interest (ROI)-based and voxelwise analyses, as well as high contrast in the images.

Protocol

NOTE: The research conducted to develop this protocol was performed in compliance with New York State Psychiatric Institute Institutional Review Board guidelines (IRB #7655). One subject was scanned for recording the protocol video, and written informed consent was obtained. Refer to the Table of Materials for details about the MRI scanner used in this protocol.

1. MRI acquisition parameters

1. Prepare to acquire high-resolution T1w images using a 3D magnetization prepared rapid acquisition gradient echo (MPRAGE) sequence with the following parameters: spatial resolution = 0.8 x 0.8 x 0.8 mm³; field-of-view (FOV) = 176 x 240 x 240 mm³; echo time (TE) = 3.43 ms; repetition time (TR) = 2462 ms; inversion time (TI) = 1060 ms; flip angle = 8°; in-plane parallel imaging factor (ARC) = 2; through-plane parallel imaging factor (ARC) = 2³³; bandwidth = 208 Hz/pixel; total acquisition time = 6 min 39 s.
2. Prepare to acquire NM-MRI images using a two-dimensional (2D) gradient recalled echo sequence with magnetization transfer contrast (2D GRE-MTC) with the following parameters: resolution = 0.43 x 0.43 mm²; FOV = 220 x 220 mm²; slice-thickness = 1.5 mm; 20 slices; slice gap = 0 mm; TE = 4.8 ms; TR = 500 ms; flip angle = 40°; bandwidth = 122 Hz/pixel; MT frequency offset = 1.2 kHz; MT pulse duration = 8 ms; MT flip angle = 670°; number of averages = 5; total acquisition time = 10 min 4 s.
   NOTE: Although the displayed results used these MRI acquisition parameters, this protocol is valid for various T1w and NM-MRI imaging protocols. The NM-MRI protocol should cover ~25 mm in the inferior-superior direction to guarantee complete coverage of the SN.

2. Placement of NM-MRI volume

1. Acquire a high-resolution T1w image (≤1 mm isotropic voxel size). Use online reformatting directly after image acquisition to create high-resolution T1w images aligned to the anterior commissure-posterior commissure (AC-PC) line and the midline.
   1. Carry out online reformatting using the vendor-provided software (e.g., if acquiring data on a GE scanner: MultiPlanar Reconstruction (MPR) in Planning; if acquiring data on a Siemens scanner: MPR in the 3D Task Card; if acquiring data on a Philips scanner: MPR in the Render Mode of the VolumeView Package).
      1. Create multiplanar reconstructions of the 3D T1w image in the axial plane perpendicular to the AC-PC line to cover the whole brain with minimal slice-gap.
      2. Create multiplanar reconstructions of the 3D T1w image in the coronal plane perpendicular to the AC-PC line to cover the whole brain with minimal slice-gap.
      3. Create multiplanar reconstructions of the 3D T1w image in the sagittal plane parallel to the AC-PC line to cover the whole brain with minimal slice-gap.
2. Load the sagittal, coronal, and axial views of the reformatted high-resolution T1w image and ensure that reference lines depicting the location of each displayed slice are present.
3. Identify the sagittal image that shows the largest separation between the midbrain and thalamus (Figure 1A). To do this, visually inspect the sagittal slices of the reformatted T1w image until the slice showing this greatest separation is identified.
4. Using the sagittal image from the end of step 2.3, visually identify the coronal plane that delineates the most anterior aspect of the midbrain (Figure 1B).
5. Using the coronal image from the end of step 2.4, visually identify the axial plane that delineates the inferior aspect of the third ventricle (Figure 1C).
6. On the sagittal image from the end of step 2.3, align the superior boundary of the NM-MRI volume to the axial plane identified in step 2.5 (Figure 1D).
7. Move the superior boundary of the NM-MRI volume 3 mm in the superior direction (Figure 1E).
8. Align the NM-MRI volume to the midline in the axial and coronal images (Figure 1F).
9. Acquire the NM-MRI images.

Figure 1: Images displaying the step-by-step NM-MRI volume placement procedure. Yellow lines indicate the location of the slices used for volume placement as described in the protocol. (A) First, the sagittal image with the greatest separation between the midbrain and thalamus is identified (step 2.3 of the protocol). (B) Second, using the image from A, the coronal plane delineating the most anterior aspect of the midbrain is identified (step 2.4). (C) Third, on the coronal image from the plane identified in B, the axial plane delineating the inferior aspect of the third ventricle is identified (step 2.5). (D) Fourth, the axial plane identified in C is displayed on the sagittal image from A (step 2.6). (E) Fifth, the axial plane from D is shifted 3 mm in the superior direction, and this plane indicates the superior boundary of the NM-MRI volume (step 2.7). (F) The final NM-MRI volume placement where the coronal image corresponds to C, the sagittal image corresponds to A, and the axial image corresponds to the axial plane in E. The NM-MRI volume is aligned to the brain midline in the coronal and axial images and the AC-PC line in the sagittal image (step 2.8). Part of this figure has been reprinted with permission from Elsevier from ³⁰. Abbreviations: NM-MRI = neuromelanin-sensitive magnetic resonance imaging; AC-PC = anterior commissure-posterior commissure. Please click here to view a larger version of this figure.

3. Quality control checks

1. Ensure that the acquired NM-MRI images cover the entire SN and that the SN is visible in the central images but not in the most superior or most inferior images of the NM-MRI volume. Otherwise (Figure 2), repeat steps 2.3-2.9 to ensure correct NM-MRI volume placement. If the participant has moved significantly since the acquisition of the high-resolution T1w scan, repeat steps 2.1-2.9.

Figure 2: Example of an NM-MRI acquisition that failed the first quality control check (step 3.1 of the protocol). Each of the 20 NM-MRI slices displayed from most inferior (top left image) to most superior (bottom right image); the image window/level was set to exaggerate the contrast between the substantia nigra and crus cerebri. The orange arrows in slices 15-19 show the location of the substantia nigra in those slices. The red arrow in the most superior slice (slice 20) shows that the substantia nigra is still visible in this slice, and thus, the acquisition fails the quality check. Abbreviation: NM-MRI = neuromelanin-sensitive magnetic resonance imaging. Please click here to view a larger version of this figure.

2. Check for artifacts, particularly ones that go through the SN and the surrounding white matter, by visually inspecting each slice of the acquired NM-MRI scan.
   1. Look for abrupt changes in signal intensity with a linear pattern that does not respect normal anatomical boundaries. For example, this may appear as a low-intensity region that is flanked by two high-intensity regions.
   2. If the artifact is the result of blood vessels (Figure 3A), retain the NM-MRI images because these artifacts will most likely always be present.
   3. If the artifacts are the result of participant head motion (Figure 3B), remind the participant to stay as still as possible and reacquire the NM-MRI images according to step 3.2.5.
   4. If the artifacts are ambiguous (Figure 3C), reacquire the NM-MRI images according to step 3.2.5. Upon reacquisition, if the artifacts remain present, proceed with these images as they are likely biological rather than a result of acquisition issues.
   5. If the NM-MRI images pass the quality control check in step 3.1, copy the previous NM-MRI volume placement. If the NM-MRI images fail the quality control check in step 3.1, repeat steps 2.3-2.9 to ensure correct NM-MRI volume placement (or steps 2.1-2.9 if the participant moved significantly).

Figure 3: Examples of NM-MRI acquisitions that failed the second quality control check (step 3.2 of the protocol). Only one representative slice is shown for each case. (A) An NM-MRI acquisition that fails the quality control check due to a blood vessel artifact (red arrows) that is the result of the blood vessel identified by the blue arrows. (B) An NM-MRI acquisition that fails the quality control check due to motion artifacts (red arrows). (C) An NM-MRI acquisition that fails the quality control check due to an ambiguous artifact (red arrows). Abbreviation: NM-MRI = neuromelanin-sensitive magnetic resonance imaging. Please click here to view a larger version of this figure.

Results

Figure 4 shows the representative results from a 28-year-old female participant with no psychiatric or neurological disorders. The NM-MRI protocol ensures complete coverage of the SN, achieved by following step 2 of the protocol outlined in Figure 1, and satisfactory NM-MRI images by following step 3 of the protocol. Excellent contrast between the SN and neighboring white matter regions with negligible NM concentration (i.e., crus cerebri) can be seen. Thes...

Discussion

The dopaminergic system plays a crucial role in healthy cognition and neuropsychiatric disorders. The development of noninvasive methods that can be used to repeatedly investigate the dopaminergic system in vivo is critical for the development of clinically meaningful biomarkers. The protocol described here supplies step-by-step instructions for acquiring good-quality NM-MRI images of the SN, including placement of the NM-MRI volume and quality control checks to ensure usable data.

Ev...

Materials

Name	Company	Catalog Number	Comments
3T Magnetic Resonance Imaging	General Electric		GE SIGNA Premier with 48-channel head coil