Cross-Platform MRI/PET or MRI/SPECT Imaging, and Co-Registration

Todd A. Sasser (1), Sarah E. Chapman (2), Ian Sanders (2), Lucas Liepert (2), W. Matthew Leevy (2,3)

Author Information: 1Bruker Biospin Inc., 44 Manning Rd, Billerica, MA, 01821 ; 2Department of Chemistry and Biochemistry, 236 Nieuwland Science Hall, University of Notre Dame, Notre Dame, IN 46556 ; 3Notre Dame Integrated Imaging Facility, University of Notre Dame, Notre Dame, IN 46556.

Application Overview

Individual in vivo imaging modalities have relative strengths (i.e. sensitivity, resolution, quantitative accuracy, throughput capabilities) for molecular, anatomical and physiological detections in preclinical and clinical oncology (Keunen et al., 2014, James & Gambhir, 2012). Fluorescence (FLI) and bioluminescence imaging (BLI), Positron Emission Tomography (PET), and Single-Photon Emission Computed Tomography (SPECT) provide functional information while X-ray Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and UltraSound (US) provide mostly anatomical information. BLI offers high throughput and highly sensitive detections (Müller et al., 2013). PET and SPECT are quantitative and can allow for dynamic imaging. Imaging techniques can be combined to leverage the strengths of multiple modalities. Most frequently, a functional imaging modality is performed in sequence with an anatomical modality to obtain an anatomical reference. Combined and system-integrated SPECT/CT and PET/CT is now common in the preclinical imaging field. There is currently a growing interest in combining functional imaging with MRI (Sauter et al., 2010). This is driven in part by a desire to avoid the specimen dose associated with CT imaging. Additionally, MRI provides superior soft tissue contrast and the ability to perform advanced MRI techniques such as diffusion weighted imaging (DWI) that can be leveraged in oncology studies (Preuss et al., 2014). Multiple functional imaging modalities may also be performed on the same sample. For example, BLI is highly sensitive and may be employed to monitor tumor development from the early stages of development. This may be combined with highly quantitative PET imaging in studies of therapeutic response. A full discussion of the merits of the various imaging modalities is available in the recent reviews by de Jong et al. (2014), Albanese et al. (2013), and James & Gambhir (2012).



For SPECT imaging, Foxn1nu mice (Jackson Laboratories; Bar Harbor, ME) were grafted subcutaneous with 700,000 HCT 116-hNIS-NEO (Imanis Life Science; Rochester, MN) tumor cells expressing the human sodium iodide symporter. Imaging was performed at 2-3 weeks post implantation when tumors reached approximately 1-3 mm diameter. Foxn1nu mice receiving no cell grafts or treatments were used for this initial validation of cross-platform 18FDG-PET/MR imaging.


PET and SPECT acquisitions were collected using the Bruker Albira PET/SPECT/CT imaging system. For PET and SPECT imaging mice received between 100 and 200 μCi 18F-FDG and 500 μCi and 1 mCi 99mTc respectively, one hour prior to imaging. MR imaging was performed using the Bruker ICON 1T MR system using a 3D MR T2-weighted RARE sequence.

Multimodal Animal Bed and Fiducial Marker Lid

The Bruker Multimodal Animal Bed (MMAB) was used for transferring animals between the Bruker Albira PET/SPECT/CT imaging system and the Bruker ICON 1T MR imaging system. The fiducial markers lid (FML) was designed with a 1 mL circuitous fill line. The line extends the length of the bed. This design provides multiple reference points for fiducial matching. The fill line was fit with threaded inlet ports to facilitate filling. The FML was printed using the Projet printer using Somos WaterClear Ultra 10122.

Image Registration

Image registration was performed using PMOD Technologies (Zurich, Switzerland) Fuse It module, provided standard with the Albira PET/SPECT/CT imaging systems. This module allows for dual image plane  adjustments including rotations, X, Y, Z movements, and image re-scaling. We used manual registration of the FML line between cross-platform images.


During initial cross-platform studies made in the absence of fiducial marker solutions we found post imaging registration to be subjective, relying largely on anatomical/physiological landmarks. Fiducial
capillary tubes filled with mixed contrast agents are frequently used to facilitate cross-platform imaging and we later employed capillary tubes filled with contrast agents. Filling and placement of such tubes and secure placement within the confines of an immobilization bed add additional complications to the study preparation process. To facilitate more streamlined fiducial matching we designed a FML (Figure 1) that is compatible with the Bruker MMABs. The FML line can be filled easily using a fill syringe and Luer adapter.

Additionally, the FML lid conforms to the general mold of the Bruker MMAB factory lid and connects similarly to the MMAB bed base (Figure 2). The FML line extends the length of the bed so it can be
utilized for any anatomical region.

Figure 1


Figure 1: FML with fill syringe installed. The FML inlets are fit with threaded Luers for ease of filling. The line may be filled with mixed contrast agents for MR (water), PET/SPECT (diluted radionuclide), and CT (radiopaque solution).

Figure 2


Figure 2: FML lid installed on MMAB base and mounted to Albira MMAB docking arm.

We initially tested the FML solution with a Foxn1nu mouse without tumor grafting or other experimental treatments (Figure 3). The circuitous FML fill line served as an ideal point of reference to register images. The numerous X, Y, and Z coordinates of the line allowed for simple image registration using the PMOD Fuse It module. Mouse 18FDG-PET/MR tumor studies employing this protocol are currently ongoing.

Figure 3

Fig3 cross platform

Figure 3: Cross-platform PET/MR image registration of healthy mouse imaged in MMAB with the FML. (A) Image registration between 18FDG-PET (fire) and MR (gray) was made using the Bruker MMABs equipped with the FML and filled with 18FDG in solution. The image panel to the left shows a coronal view plane with the straight region (arrow) of the FML fill line used to register images. The image panel to the right shows a coronal view plane with the circuitous region (arrow) of the FML fill line used to register images. (B) Four slice sequence of PET/MR coronal view of hind region of mouse with kidney (lower arrow left) and spine (upper arrow left) 18FDG signal apparent.

We next evaluated the protocol for cross-platform 99mTc-SPECT/MR imaging in a HCT 116-hNIS-NEO tumor model. This imaging protocol and registration method resulted in excellent SPECT and MR tumor signal/contrast registration (Figure 4).

Figure 4


Figure 4: Cross-platform 99mTc-SPECT (rainbow) and MR (gray) imaging of HCT 116-hNIS-NEO tumor mouse with image registration facilitated by the MMAB with the FML. Top row is MR only in transverse, sagittal and coronal views. Middle row is SPECT only in transverse, sagittal and coronal views. Bottom row is registered MR/SPECT in transverse, sagittal and coronal views.

Frequently, preclinical researchers are required to generate VOIs for tumor analysis based only on the functional PET/SPECT images, either because the integrated CT imaging does not provide sufficient tumor contrast and/or because cross-platform CT/MR image registration is not suitably accurate. However, functional modalities will not always accurately show the true tumor margin, and ideally users would have an option to generate VOIs based on an anatomical visualization of the tumor margin. Our results provide for excellent tumor margin contrast with MRI and reliable cross-platform PET or SPECT registration. This protocol should allow for accurate production of VOIs based on tumor margins identified in MR images and application to functional PET or SPECT images.


Individual functional and structural imaging modalities can be combined for enhanced analytical value. There is a growing trend of combining PET and MR imaging. To fully leverage the potential of cross-platform PET/MR imaging, methods to ensure animal stability and image registration should be considered. Here we describe a method using a commercial Bruker Multimodal Animal Bed with a custom Fiducial Marker Lid for performing cross-platform PET or SPECT with MR that will allow for reliable transport and image registration. The FML fill line is suitable for PET, SPECT, CT, and MR contrast agents and requires minimal preparation, relative to alternative solutions that use makeshift beds and/or capillary tubes, to be used in imaging studies.

→ Read the References

→ Download the entire PDF

Learn more