ACTM: Adaptive Computed Tomography With Modulated-Energy X-Ray Pulses

X-ray Computed Tomography (CT) is a widely used tool for security and industrial non-destructive testing (NDT). Scanned objects in these uses typically contain multiple materials with different atomic numbers. In these cases, Dual-Energy CT (DECT) has numerous advantages over conventional single-energy CT, including improved image quality, reduction of effect of scatter and other artefacts, and better material decomposition. Therefore, DECT has become the standard for inspection of objects containing multiple materials [1-2]. In this paper, we present a new CT technology: Adaptive Computed Tomography with Modulated-energy X-ray pulses (ACTM) [3]. ACTM is based on adaptive cargo radiography techniques [4-6] that have been developed over the last few years. To perform CT imaging, we propose to utilize the exact and fast Katsevich image reconstruction algorithm [7-12].Key enabling ACTM techniques are: · Linear accelerator with deep energy control. · Multi-energy spectrum in each slice provided by packets of short X-ray pulses with end-point energy ramp up (> 1 MHz rate of energy switching). · Fast detectors with reduced sensitivity to scatter radiation. · Detector readout with silicon photomultipliers (SiPM) provides time-resolved measurement of short X-ray pulses. · Electronics and an algorithm for automatic, dynamic adjustment of SiPM responsivity in detector channels provides increased dynamic range of X-ray imaging. · Real-time algorithm for X-ray pulse-by-pulse selection of high and low energy windows for dual-energy material decomposition. · Filtered Backprojection (FBP) helical image reconstruction based on the theoretically exact Katsevich algorithm.The ACTM method for dual- (or, multi-) energy CT mitigates the main disadvantages of the conventional interlaced approach: · Ambiguity and artifacts caused by sampling different regions due to motion of the object between interlaced pulses with distinct energies. · Distortion of CT image of the boundaries between regions with large difference in Z. · Small range of object thicknesses where material decomposition is valid.