Abstract: A Compressed Sensing (CS) based image reconstruction method and system is described herein which may be used to reduce the X-ray dose radiation in Computed Tomography (CT) or to decrease the scan duration in MR imaging (MRI). Methods and systems described herein may address problems that have hindered the clinical usage of CS, i.e. computation complexity and modeling problems. Using the described algorithm, high quality images may be recovered from undersampled data which may help to reduce the scan time and the exposed invasive radiations. Using the same set of data in conventional image reconstruction algorithms (e.g. Filtered Back Projection (FBP) in CT) may cause severe streak artifacts and may take significantly more time using Graphics Processing Units (GPU) and parallel clusters with the conventional CS-based methods. This method can be used other imaging modalities using Radon transform (such as C-Arm and electron tomography, for example).

Abstract: A Compressed Sensing (CS) based image reconstruction method and system is described herein which may be used to reduce the X-ray dose radiation in Computed Tomography (CT) or to decrease the scan duration in MR imaging (MRI). Methods and systems described herein may address problems that have hindered the clinical usage of CS, i.e. computation complexity and modeling problems. Using the described algorithm, high quality images may be recovered from undersampled data which may help to reduce the scan time and the exposed invasive radiations. Using the same set of data in conventional image reconstruction algorithms (e.g. Filtered Back Projection (FBP) in CT) may cause severe streak artifacts and may take significantly more time using Graphics Processing Units (GPU) and parallel clusters with the conventional CS-based methods. This method can be used other imaging modalities using Radon transform (such as C-Arm and electron tomography, for example).