Data traffic compression in spectral photon-counting CT imaging based on human visual characteristics

Abstract

Spectral computed tomography (CT) with photon-counting detectors (PCDs) can provide a variety of cross-sectional images, showing improved diagnostic capability at low radiation doses. PCDs acquire and divide signals of each energy intensity into separate energy bins and improve the count rate by pixel size miniaturization, which can simultaneously increase spatial resolution. Nevertheless, PCDs do show some limitations in data processing. Acquisition of large-spectrum data requires large spaces of an analog-to-digital converter (ADC) and a memory in each pixel, and this process generates massive amounts of data traffic. The conventional compression techniques used in CT imaging are not suitable for such traffic because they are applied to the CT image data, not the projection data stored by PCDs. This study aimed to secure ADC and memory space and compress the data traffic while maintaining high image resolution by using a number of energy bin's information for computing. This approach was based on a composite method using CT images from high spatial resolution non-spectral CT and low spatial resolution spectral CT. The resolution in non-spectral CT was recognized by assigning grayscale values, while the resolution for spectral CT was maintained with a pseudocolor scale based on visual characteristics. In this paper, we evaluated the image quality against the compression rate to verify our concept.