Low-dose graphic-processing-unit based limited-angle CT reconstruction algorithm development for a home-designed dual modality micro-FT/CT system

Computed tomography (CT) has evolved into an indispensable imaging method in clinical routine for a long time. It gives us high resolution, high contrast and non-invasive tomographic images in a brief scan time. However, their high radiation dose is an important issue to be solved in CT system development. Iterative algorithms can give us better image quality than analytical types and reduce some artifact effectively. However, they are not only time consuming, but also need very large-scale memory size. In this study, we reduce our projection data to <180-degree limited angle reconstruction, which was based on compressive sensing (CS) criteria and total variation (TV) minimization. This is a computer simulation study combined with CPU and graphic processing units (GPUs). We developed a 3D system matrix in which ray-tracing and circular redundancy technique were included. All CT image data were collected by our homemade micro- FT/CT dual modality system and reconstructed by model based post total-variation constrained algebraic reconstruction technique (ART-TV) algorithm, inner total-variation constrained expectation-maximization (TV-EM) and post total-variation constrained expectation- maximization (EM-TV) in transmission types. The reconstruction volume was 216 × 216 × 384 by 4 × 4 binning of the detector size. There is only 90-degree angular sampling range with 1 degree/projection. Some figures of merit (FOMs) such as rootmean- square error (RMSE), universal quality index (UQI) and coefficient of variation (CV) were used to evaluate image quality. Although now the total speed up ratio is about 1.3 × than CPU multi-threading but compare to our previous study, we push our reconstruction scale from simulated 16 ×16 × 16 and 32 × 34 detector size to the size of a realistic scan. Our study demonstrates that statistic reconstruction technique (SRT) with < 180-degree limited angle reconstruction can produce low-dose CT images.