Click here to


Are you sure ?

Yes, do it No, cancel

Characterization of Bioluminescence Tomography-Guided System (MuriGlo) for Pre-Clinical Radiation Research

X Xu1*, Z Deng1 , P Tsouchlos2 , T Kanesalingam2 , H Dehghani3 , J Wong1 , K Wang1 , (1) Johns Hopkins School of Medicine, Baltimore, MD, (2)Xstrahl Inc. ,Suwanee,Georgia, (3) School of Computer Science, University of Birmingham, West midlands, UK


(Wednesday, 7/17/2019) 4:30 PM - 6:00 PM

Room: 304ABC

Purpose: To advance image-guided irradiation for soft-tissue targeting, the Hopkins team has translated the technology to Xstrahl Inc to develop a 3D bioluminescence tomography(BLT, MuriGlo) for small animal radiation research platform(SARRP). We characterized the MuriGlo system and demonstrated its capability to localize soft tissue targets, which are usually challenging for CBCT to localize.

Methods: Four-mirror system, filters, lens and CCD camera together formed a compact design to allow a desktop size imaging platform and to support multi-projection and multi-spectral imaging for superior BLT localization(Fig. 1). The moveable mouse bed allows animal imaged in MuriGlo and transferred to SARRP for CBCT and irradiation. A geometrical registration based on pinhole camera model was applied to map 2D BL images(BLIs) to the animal surface generated from the CBCT image for BLT reconstruction. Unfiltered optical images were taken before and after filtered images to record the in vivo signal change and correct the filtered images for optical reconstruction. As the animal curvature relative to the detector/camera position can affect the surface bioluminescent signal used for reconstruction, to effectively eliminate this effect, we have implemented a novel spectral derivative method with compressed sensing conjugate gradient algorithm. Phantom and glioblastoma(GBM) mouse model are used to validate the localization accuracy of the optical system.

Results: The 2D BLIs mapped to the CBCT image is at average accuracy of 0.3 mm(Fig. 2). In vivo signal variation was characterized to correct the contribution of the multi-spectral images taken at different time points(Fig. 3). Our initial result shows that MuriGlo can qualitatively retrieve target shape within 1mm localization accuracy by comparing the center of mass between BLT and contrast-labelled CBCT target volume for both phantom(Fig. 4) and in vivo GBM mouse model(Fig. 5).

Conclusion: The MuriGlo system is expected to provide a new quantitative imaging guidance for pre-clinical radiation research.

Funding Support, Disclosures, and Conflict of Interest: Xstrahl Sponser research agreement: 90043185 Dr. Wong receives royalty payment from a licensing agreement between Xstrahl Ltd. and Johns Hopkins University


Optical Tomography, Image-guided Therapy, Quantitative Imaging


TH- Small Animal RT: General (most aspects)

Contact Email