Multiresolution iterative reconstruction for region of interest imaging in X-ray cone-beam computed tomography

#This disclosure relates to reconstructing images in computed tomography (CT) using various regions of different resolution in the image domain and corresponding regions of different resolution in the sinogram/projection domain, and, more particularly, to selecting the regions of high resolution in the image and sinogram/projection domains to correspond with regions of high spatial frequencies and/or regions identified to have clinical significance.

#Computed tomography (CT) systems and methods are widely used, particularly for medical imaging and diagnosis. CT systems generally create images of one or more sectional slices through a subject's body. A radiation source, such as an X-ray source, irradiates the body from one side. At least one detector on the opposite side of the body receives radiation transmitted through the body. The attenuation of the radiation transmitted through the body is measured by processing electrical signals received from the detector.

#A CT sinogram indicates attenuation through the body as a function of position along a detector array and as a function of the projection angle between the X-ray source and the detector array for various projection measurements. In a sinogram, the spatial dimensions refer to the position along the array of X-ray detectors. The time/angle dimension refers to the projection angle of X-rays, which changes as a function of time during a CT scan. The attenuation resulting from a portion of the imaged object (e.g., a vertebra) will trace out a sine wave around the vertical axis. Those portions farther from the axis of rotation correspond to sine waves with larger amplitudes, and the phases of the sine waves correspond to the angular positions of objects around the rotation axis. Performing an inverse Radon transform—or any other image reconstruction method—reconstructs an image from the projection data in the sinogram.

#A more complete understanding of this disclosure is provided by reference to the following detailed description when considered in connection with the accompanying drawings, wherein:

#FIG. 1 shows a flow diagram of a first implementation of a multiresolution iterative reconstruction (IR) method, according to one implementation;

#FIG. 2 shows high-resolution and low-resolution regions within projection data and within a reconstructed image generated from the projection data, according to one implementation;

#In computed tomography (CT) and cone-beam CT (CBCT), region of interest (ROI) reconstruction is used to obtain high-resolution images within a designated ROI. This technique can also be referred to as zoomed ROI reconstruction, and uses a smaller image voxel size for representation of anatomic features in the designated ROI. Further, in certain implementations, ROI reconstruction can also generate coarse-resolution images and information for regions outside the ROI. The methods and apparatus described herein provide computational and other advantages over more conventional methods by, among other things, using different pixel pitches for the projection data corresponding to regions outside the ROI relative to the pixel pitch used for regions projecting through the ROI. In contrast, more conventional methods do not use multiple pixel pitches within the projection data for a single CT scan. Further, methods described herein perform ROI reconstruction using iterative reconstruction based on a single scan, rather than multiple scans such as are used in a two-pass method.

#Often for CT using iterative reconstruction (IR), the entire imaged object OBJ is within the image volume of the CT scanner. This reduces the truncation effect. However, when high-resolution images are being reconstructed, decreasing the imaged region to a small ROI can reduce the computational burden for reconstructing an image of the object OBJ. Otherwise the time required to reconstruct a high-quality image can become prohibitive, especially when high-resolution reconstruction for the ROI is desired, and the same high-resolution is used for the entire volume of the reconstructed image.

#The computational time for iterative image reconstruction is often dominated by the forward-projection operations and back-projection operations. This is because often IR algorithms converge to the reconstructed image using multiple steps of iteration, and each step can include both a forward-projection operation and a back-projection operation. The number of floating point operations for each of these operations can be of the order of the number of voxels in the image domain times the number of pixels in the projection/sinogram domain. As used herein, the terms “sinogram domain” and “projection domain” are used interchangeable. Thus, doubling the resolution without changing the size of the image or projection domains can result in a 25=32 times increase in the computational time for each forward- and back-projection operation.