Translator Disclaimer
9 March 2017 Reconstruction method for x-ray imaging capsule
Author Affiliations +
A colon imaging capsule has been developed by Check-Cap Ltd (C-Scan® Cap). For the procedure, the patient swallows a small amount of standard iodinated contrast agent. To create images, three rotating X-ray beams are emitted towards the colon wall. Part of the X-ray photons are backscattered from the contrast medium and the colon. These photons are collected by an omnidirectional array of energy discriminating photon counting detectors (CdTe/CZT) within the capsule. X-ray fluorescence (XRF) and Compton backscattering photons pertain different energies and are counted separately by the detection electronics. The current work examines a new statistical approach for the algorithm that reconstructs the lining of the colon wall from the X-ray detector readings. The algorithm performs numerical optimization for finding the solution to the inverse problem applied to a physical forward model, reflecting the behavior of the system. The forward model that was employed, accounts for the following major factors: the two mechanisms of dependence between the distance to the colon wall and the number photons, directional scatter distributions, and relative orientations between beams and detectors. A calibration procedure has been put in place to adjust the coefficients of the forward model for the specific capsule geometry, radiation source characteristics, and the detector response. The performance of the algorithm was examined in phantom experiments and demonstrated high correlation between actual phantom shape and x-ray image reconstruction. Evaluation is underway to assess the algorithm performance in clinical setting.
© (2017) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Daniel Rubin, Ronen Lifshitz, Omer Bar-Ilan, Noam Weiss, Yoel Shapiro, and Yoav Kimchy "Reconstruction method for x-ray imaging capsule", Proc. SPIE 10132, Medical Imaging 2017: Physics of Medical Imaging, 101324S (9 March 2017);

Back to Top