Erasmus MC : High Resolution MRI and Translational Research

Deze pagina is nog niet in het Nederlands beschikbaar.

Principal Investigator for Radiology:   Piotr A Wielopolski, PhD

This research line is dedicated to making translational research accessible to clinicians and researchers through a setup of optimized tools for imaging at all resolution ranges on clinical imaging platforms.  We seek to improve throughput and resolution of 1.5T and 3.0T clinical MRI scanners both for in vitro and in vivo in order to facilitate large-scale trials of molecular imaging techniques   These tools cover distinct areas, such as multi-channel signal reception and dedicated receiver coils (in collaboration with Herman Flick, Flick Engineering Solutions BV, the Netherlands), MRI pulse sequence development and support, and measurement protocol improvements (in collaboration with Gavin Houston, General Electric Healthcare, Applied Science Laboratory Europe).  This research line also intends to expand ongoing projects in our department and, most importantly, creates a collaboration platform so that a multi-disciplinary approach can be instantiated when new methodologies are available to facilitate communication channels between different imaging modalities.

MR micro- and molecular imaging: Phased array and parallel imaging technology on whole-body unmodified clinical scanners

This project investigates the micro-imaging potential of clinical whole-body MRI scanners to provide a flexible platform for translational research and molecular imaging.  Specially developed 4 and 8 phased array channel connection interfaces (manifolds) and custom built surface coils make our 1.5T and 3.0T whole-body clinical imaging scanners well suited for this purpose.  The typical clinical gradient hardware (a nominal strength of 40 mT/m - 260 ms rise time to gradient maximum) in combination with tailored MRI pulse sequences has been adapted to perform imaging at micron resolution (10 - 100 µm2  pixels) (Figure 1).  The phased array interface makes it possible to extend the field-of-view to suit particular applications ranging from continuous larger area scanning (e.g. 5 x 5 cm2) to multi-local imaging in different samples with good signal- and contrast-to-noise ratios (SNR/CNR) (Figure 2).  Combined pair loop and quadrature coil arrangements without interaction between coils in the direction perpendicular to the coils to further improve SNR.  Parallel imaging is possible with this multi-channel interface, making it possible to include echo planar imaging (EPI) protocols for small animal imaging (Figure 3).  EPI can provide can provide fast sampling of different contrasts for more precise quantification, e.g. to monitor contrast uptake in small tumors at reasonable good voxel resolution of 0.7-1.0 x 0.5 x 0.5 mm3, sub-seconds data collection and acceptable geometrical distortions.

Figure 1: High resolution MRI on a 3.0T clinical scanner.   Specialized custom designed multi-channel reception technology (up to 8 channels) and small surface coils can provide the necessary signal-to-noise and large volume coverage necessary for high resolution imaging applications for translational research on clinical scanner platforms.  Above, a transverse section of a female rat pelvis demonstrates the urinary sphincter (A) and a magnified view (B).  Below, two apple seeds are imaged at the highest in-plane resolution presently possible (9.8 x 9.8 µm2) with a T2*-weighted contrast, a contrast setting that is adequate for tracking SPIO labeled cells in vivo (C,D).

Figure 2: Speeding data collection time in multiple samples with high resolution and SNR using phased array imaging technology.  To streamline studies in animals undergoing the same imaging protocol, a multiple receiver channel interface makes it possible to scan in parallel from 2-8 regions of interest in parallel with high signal-to-noise ratios both at 1.5T and 3.0T using independent coils and a 4 or 8 reception channel interface.  This example demonstrates two high resolution angiograms (MIP) of sarcoma tumors in the hind limbs of two rodents undergoing an ILP (isolated limp perfusion) therapy after the administration of an intravascular contrast agent collected at 1.5T collected in 15 minutes during the same imaging session.

Figure 3: High resolution echo planar imaging (EPI) of a rat heart on a 1.5T MRI clinical platform.  The possibilities with a clinical scanner gradient hardware are explored for higher resolution EPI at 1.5T. A rat heart is demonstrated with different contrast weightings.  A standard T1-weighted 3D scan at high resolution is compared to a proton density-weighted (B) and mildly T2-weighted (C) EPI scans with a voxel size of 0.7 x 0.5 x 0.6 mm3. Two surface coils were used for imaging.  

High resolution MRI using susceptibility contrast for the detection of vascular structures

A contrast setting that is becoming increasingly interesting in the clinical arena is that provided by sequences that enhance the magnetic susceptibility differences between tissues, providing improved visualization of particular structures.  Magnetic susceptibility effects are best seen using T2*-weighted gradient echo (GRE) sequences.  These effects increase at higher magnetic field strengths so applications using this type of contrast are best suited on our 3.0T as the T2* shortening is stronger at shorter echo times (TEs) as compared to lower field strengths (e.g. 1.5T).  With this contrast, venous structures and blood by-products can be nicely demonstrated.  This is illustrated in a high resolution image of a sarcoma implanted in the limb of rat where the lower signals observed in the tumor periphery correspond to veins. T2*-weighted contrast also provides an effective mean of tracking blood by-products, such as hemosiderin in micro-hemorrhages and chronic conditions (such as in endometriosis).

Figure 4:  Vascular architecture in tumors using susceptibility contrast at 1.5T.  T2*-weighted 3D scans using a long TE can display the venous vascular structure effectively.  The appearance of hypointense areas in the periphery of the sarcoma (arrows) illustrates venous structures and points at the richer blood supply that sustains the tumor growth. Higher magnetic field strength, e.g. 3.0T, will enhance the effect of magnetic susceptibility with short TEs.

High resolution vascular imaging

New blood vessel formation is essential to the unrestricted growth of tumors.  Angiogenesis is the process by which new vessels form in both benign and malignant processes.  The potential of a tumor to successfully metastasize can be linked to angiogenesis. More angiogenic tumors possess a greater number of microvessels through which the malignant cells are shed into the blood stream and metastasize.  Imaging the vasculature surrounding tumors is of great interest to us and provides a great collaborative channel with other research departments (Cardiology, Nuclear Medicine, Surgical Oncology and Pathology).  Several MRI blood pool contrast agents have been recently approved for clinical use, making it possible to image the vascular tree around the tumor non-invasively with high SNR and at high resolution.  Presently imaging is performed in tumor models such as the skinfold window model (Figure 5).  Short TE T1-weigthed 3D imaging has been possible with our research clinical scanners at resolutions ranging between 60 - 100 µm2   pixels and 100 µm slices.  Results also show that the assessment of tumor angiogenesis could be performed more accurately using blood pool contrast agents than with the widely available low molecular contrast agents that diffuse rapidly through the capillary walls of both normal and tumor vessels.  These developments open the potential to monitoring tumor response more adequately and translate directly the results obtained from therapeutic trials in animal models to clinical reality.  

Figure 5:  High resolution vascular imaging using intravascular contrast agents.  Intravascular contrast agents and a dedicated 1 cm ID coil at 3.0T can provide exquisite detail of the vascular signal surrounding a developing tumor (sarcoma, arrows) in a mouse magnetically compatible skinfold window model.  The surface coil fits precisely over the glass, providing the high signal necessary for imaging voxels on the order of 70 x 70 x 100 µm3 with scan times below 20 minutes.  The thickness of the window containing the tumor is approximately 1.2 mm.  Gd-DTPA-Albumin or Gd-liposomes are used for enhancing the blood signal.

High resolution imaging for monitoring proliferation and trafficking of magnetically labeled stem cells

Present clinical 1.5T and 3.0T systems are particularly well suited for high-resolution tracking of positively and negatively labeled stem cells for direct translational imaging for molecular imaging applications in the clinics.  Experiments performed in animal MRI systems acquired at higher magnetic field strengths may be difficult to translate as inherent changes in the properties of tissues and contrast enhancing agents can change dramatically.  The 3.0T scanner could provide the optimal setup regarding operational field strength with enough T2*-weighted contrast for tracking iron-loaded cells or T1-weighted contrast with short TE for vascular or positively enhanced labeled cells (e.g. using Gd-DTPA bearing liposomes or Mn-DPDP).  Whole body human MRI scanners at higher magnetic field strengths, exemplified by the recent installed base of 7.0 T scanners worldwide, can prove extremely difficult to work with in these applications given the problems related to B1 field inhomogeneities and radiofrequency power deposition (tissue heating, high specific absorption ratios).  Using clinical magnets makes our setup competitive to specialized animal systems at higher fields where the availability of contrast agents for clinical applications has not yet proven effective (and non-toxic).   The large bore afforded in clinical scanners provides an effective platform for imaging many samples in parallel with multi-channel phased array technology.  Presently, scanning times at 3.0T are restricted to around 30 minutes for the highest resolution setting with 3D gradient echo sequences.  Special setups for animal handling have been devised for different experiments that, in conjunction with our specialized multi-channel reception hardware and coil sets (ranging from 1 – 3 cm ID loops) provide the sufficient SNR and large area coverage.  Parallel imaging has been incorporated to double or triple data collection speed when multiple reception channels are used. In our view, optimal coil setups are relatively inexpensive to build / or acquire in comparison with specialized animal systems at higher magnetic fields (e.g. 4.6T - 11T).

High resolution 3D gradient echo sequences and different contrast settings that enhance the presence of labeled cells or target tissues are used.   Tracking labels can be visualized using the negative enhancement of tagged structures that have been labeled with paramagnetic species (larger R2 relaxivity and setting longer echo times, TE, to improve contrast).   For high resolution T2*-weighted imaging, single SPIO labeled cells have been imaged with a 1 cm ID loop with a maximum resolution of 20 µm2 pixels - 100 µm slices under 30 minutes.   Rodents receiving iron labeled tumor cells in several locations can be imaged with a wide FOV of 9 cm and 10242 imaging matrices (Figure 6).  Cells labeled with T1-shortening contrast agents require the use of T1-weighted scans with high resolution and short TEs.   The example of positive enhancement with high resolutions (Figure 2 and Figure 6) can be readily obtained using T1-weighted 3D scans after the administration of intravascular T1-shortening agents (high R1).  Similar results are hoped for positively labeled cells in the near future.  

Figure 6:  Development over time of iron-labeled tumor cells in a clinical 3.0T scanner.  The development of iron-labeled tumor cells injected under the skin in the limb of a rat are tracked over 4 weeks.  A magnified view of the injection site (day 1) and subsequent development of the signal intensities in the tumor is shown.   A high resolution angiogram was performed at day 28 to visualize the new vascular architecture developed around the tumor.  For tracking iron-labeled cells a high resolution T2*-weighted 3D scan is used over a wide FOV (9 cm) and 4 2 cm ID surface coils with a scan time under 18 min a voxel resolution of 90 x 90 x 100 µm3.  A T1-weighted 3D scan at the same resolution is performed to enhance the vascular signal after the injection of a Gd-DTPA-albumin intravascular contrast agent.  

High resolution imaging for cartilage and monitoring of cartilage repair therapies

In the Netherlands over half a million persons suffer from osteoarthritis in their lower extremities and from these patients more than 10,000 receive a knee implants each year in order to relieve pain and improve function.  The high spatial resolution that can be achieved with MRI in addition to its exquisite contrast flexibility and non-invasiveness promotes this imaging modality as the most promising method for imaging and for evaluating the status of cartilage. To describe the status of cartilage and monitor changes in the status, we started exploring the resolution capabilities of our clinical scanners for high resolution animal work, especially at 3.0T.  

Figure 7: High resolution MRI of mouse joint and implanted SPIO-labeled stem cells in equine cartilage.  Stem cell therapy is being sought to repair damaged cartilage in several models Study of a knee joint in mice under 10 minutes with a voxel size of 60 x 90 x 500 mm3 using a surface coil of 2 cm ID.  A sagittal (A), coronal (B) and axial (C) views are demonstrated.  A scaffold with iron-labeled condrocites (green arrows) is implanted in equine cartilage that had been previously damaged (D,E) with the vision that the faith of the cells and cartilage repair followed over a long period of time. Healthy cartilage is the white structure at the edges of bone (white arrow)  

Rodent cardiac imaging in a clinical platform

Microimaging on unmodified clinical scanners with signal reception hardware adjusted to the size of the imaging target is realistic. Our initial experience proves adequate even for scanning mouse hearts. This warrants a greater effort and flexibility for translational research in molecular imaging at relevant clinical MRI field strengths.

In-plane resolution has been possible between 80-150 µm2, with 0.7-3.0 mm sections with adequate temporal resolution for heart rates of approximately 350 beats per minute with the cardiac MRI pulse sequences present in the clinical scanner.  For rapid scouting, temporal resolutions have been set to a maximum of 30/60 ms per frame for cine/ non-cine scans, respectively. For cine scans, flow compensation was selected. Gd-DTPA and intravascular contrast agent (Gd-DTPA-Albumin) have been employed to improve blood/myocardium contrast for cine scans and/or to delineate myocardial infarctions.

By placing standard cardiac electrodes on the paws of the animals all the relevant pulse sequences such as proton density- and T2-weighted black blood fast spin echo (BBFSE), standard cine, cine 1D and 2D tagging and delayed contrast enhanced scans were obtained  (Figure 8).  ECG gating hardware handled heart rates up to the maximum observed of 350 beats/minute in mice with the injected anesthetics. Scanning times are kept reasonably short (depending on the reception coil and sequence selection) and did not exceed 8 minutes per slice at the highest resolution attempted (0.9 mm sections, in-plane resolution of 80 µm2).  1D and 2D tagging could be performed with a minimum tag line thickness of 130 µm and tag line separations down to 400 µm, adequate for study these rodent hearts with the present setup.

Figure 8:  Rodent fast cardiac MRI for imaging myocardial infarction and recovery process with potential therapies.  Functional and anatomical scanning can be performed effectively in clinical scanners without any hardware and minor software modifications but the inclusion of specialized surface coils for high SNR and acquisition throughput. Scan time per slice for any of the contrast settings explored is around 2 minutes with a heart rate of 250 beats per minute. A four chamber view is planned (A) and several slices positioned to image the left ventricle in a short-axis view (B).  A T2-wegithed black blood fast spin echo scan is performed (C) to visualize the region where myocardial infarction is expected from the lack of contraction in (A,B).  Using a delayed contrast enhanced scan after the injection of Gd-DPTA, a bright infarct can be seen over a darkened myocardial signal (D).  The arrow indicates the infarcted region.

MRI of hemorrhage in the heart: Tracking of iron labeled cells

Over the past decade, transplantation of (genetically modified) stem or progenitor cells has been proposed as a revolutionary new technique for the treatment of malfunctional endogenous cell populations. An interesting example is the introduction of cells (myoblasts) that can restore damaged or dysfunctional myocardium to functionality. Cells need to be fully followed in vivo to assure accurate biodistribution, to track fate, and to correlate cell presence with outcomes. Yet, tracking is severely limited by the inability to track cells deep within the body in a noninvasive manner.

Using MRI we successfully tracked cells labeled with superparamagnetic iron-oxide particles (SPIO), as they provide signal voids that can be easily recognized. SPIO-labeled cells have been proposed as a tool to study cell engraftment after transplantation into myocardial infarcts. Reperfusion of the infarct territory, as pursued in clinical practice by rapid mechanical or thrombolytic reperfusion, can cause microvascular obstruction (MVO) and hemorrhage.

We are studying the time course of myocardial infarction to determine if blood by-products generate signal voids that would make iron-labeled cell tracking possible in practice (Figure 9). We use pigs undergoing balloon occlusion of the left circumflex artery for 2 hours followed by reperfusion. At 1 week an area of MVO was observed in 85% of animals. Furthermore, hemorrhage caused hypoenhancement on the gradient echo T2* weighted in vivo scans with a size of 26±7% of the infarct area. Hemorrhage size strongly correlated with the area of MVO (P=0.002). At 4 weeks MVO was observed in 40% of animals. Hemorrhage could still be identified using T2* weighted scanning. Ex vivo scanning mainly showed intracellular methemoglobin at 1 week (bright on T1, dark on T2), surrounded by a hemosiderin ring (dark T2, slightly dark T1). At 4 weeks, we observed a diffuse pattern of hemosiderin deposits throughout the infarct interspersed by extracellular methemoglobin and hemichromes (bright on T2 and bright or isointense on T1). Histological analysis confirmed these findings. MRI appearance of intracardiac hemorrhage follows brain hemorrhage patterns. The chronic presence of hemoglobin degradation products still makes tracking of iron-labeled cells within reperfused infarct areas a difficult task.  Injections of less than 100.000 SPIO labeled stem cells have been detected in rat heats in healthy myocardium (Figure 9).  Future work using MRI histology using different contrast settings at 1.5T and 3.0T, similarly as in the porcine model of myocardial infarction) will be pursued in the infarcted rat hearts to validate the appearance of infarct and possibilities for labeled cell tracking.

Figure 9: MRI histology in a porcine myocardial infarction model and SPIO labeled cell injection in a rat heart.  MRI is well known for all its non-invasiveness and high tissue contrast possible with different measurement sequences.  An example is shown using high resolution scans at 1.5T illustrating different contrast weightings (proton density, T1, T2 and T2*) on a pig heart (A) Even after five weeks after inducing myocardial infarction it is still possible to identify different appearances that can be expected during the oxidation of hemoglobin. SPIO labeled stems cells (both using positive and/or negative contrast imaging sequences) may be difficult to track under these conditions.  100.000 SPIO labeled stem cells injected into healthy myocardium in a rat model (arrow) with negative contrast scanned with a proton density–T2*weighted setting at 3.0T (B,C).

Fiber tracking in healthy and infarcted porcine hearts

This project is an initiative to find an alternative way of assessing cardiac function and regeneration after a myocardial infarction through analysis of the structure of the myocardium fibers. Visualization of the myocardial fibers has been recently possible in vitro and in-situ with the same technique that has been used for mapping the white matter tracks in the brain using diffusion tensor imaging (DTI).  In the event of a myocardial infarction, the progression of fiber arrangement could be attempted in-vivo to study remodeling patterns. Eventually with the application of promising therapies such as the incorporation of stem cells into the infarcted region, a better understanding on how cardiac function may be improved by monitoring the creation of healthy myocardial tissue.

Figure 10: Diffusion tensor imaging (DTI) in the heart.  Myocardial fibers can be tracked similarly as demonstrated in the brain using DTI.  A region of interest is placed in the myocardium and tailored software is used to track the fibers.  The tracked pattern is reminiscent to the spiral pattern seen during histology.  In the case of myocardial infarction, the fiber bundles stop at the region of the infarct (arrow) . (A) Illustrates a false color functional anisotropy (FA) map one a healthy porcine heart where a region of interest has been drawn around the myocardium. Different track bundles can be obtained depending on the tracking parameters used (B,C).  In the case of a 5 week old infarct, tracking of the fiber bundles does not go beyond the infarcted tissue (D).  (A-C) acquired on a fresh, in-situ heart using 25 diffusion tensor directions.  (D) was acquired in a heart in formaldehyde using 6 diffusion tensor directions.

High resolution imaging of atherosclerotic plaque

This research project supports the research line “Tissue Characterization of Atherosclerotic Plaque” of Aad van der Lugt.  Magnetic resonance imaging has a great potential to provide high resolution imaging of vessel wall for the detection and estimation of atherosclerotic plaque burden.  The best contrast possible for the detection of atherosclerotic plaque and characterization of the different plaque components has been carried out with proton density-, T2- and T1-weighted contrasts and a black-blood image sequence readout that can provide good contrast between vessel wall and lumen.   Diffusion weighted imaging and usage of intravascular contrast agents is being explored as alternative contrast mechanisms that can aid in the classification of vulnerable plaque components.

Figure 11: High resolution MRI of atherosclerotic plaque at 3.0T.  To study atherosclerotic plaque components more effectively, high resolution MR histology is acquired in specimens extracted from patients undergoing endartectomy in the carotid arteries. Several contrast settings are being studied, such as proton density (A), T2-weighted (B) and T1-weighted (C).  Diffusion weighted imaging is also under scrutiny, both in excised specimens as a potential contrast for detecting vulnerable plaque components.

Participating researchers Dept. of Radiology:  Monique Bernsen, Gavin Houston,  Kees van Dijke, Jifke Veenland, Annick Weustink, Robert Jan van Geuns, Sandra van Tiel, Marion van Vliet

Collaborations: