Cycloidal CT with CNN-based sinogram completion and in-scan generation of training data

Abstract

In x-ray computed tomography (CT), the achievable image resolution is typically limited by several pre-fixed characteristics of the x-ray source and detector. Structuring the x-ray beam using a mask with alternating opaque and transmitting septa can overcome this limit. However, the use of a mask imposes an undersampling problem: to obtain complete datasets, significant lateral sample stepping is needed in addition to the sample rotation, resulting in high x-ray doses and long acquisition times. Cycloidal CT, an alternative scanning scheme by which the sample is rotated and translated simultaneously, can provide high aperture-driven resolution without sample stepping, resulting in a lower radiation dose and faster scans. However, cycloidal sinograms are incomplete and must be restored before tomographic images can be computed. In this work, we demonstrate that high-quality images can be reconstructed by applying the recently proposed Mixed Scale Dense (MS-D) convolutional neural network (CNN) to this task. We also propose a novel training approach by which training data are acquired as part of each scan, thus removing the need for large sets of pre-existing reference data, the acquisition of which is often not practicable or possible. We present results for both simulated datasets and real-world data, showing that the combination of cycloidal CT and machine learning-based data recovery can lead to accurate high-resolution images at a limited dose.

Data-Driven Interpolation for Super-Scarce X-Ray Computed Tomography

“…For instance, [1] uses an initial bicubic interpolation to estimate missing data in the sinogram, which is then further improved using a U-net [2] deep neural network, whilst [3] uses a linear interpolation followed by a U-net. [4] also uses bicubic spline interpolation and convolutional neural network enhancement, whilst [5] uses bilinear interpolation followed by a regularisation performed during reconstruction, using the regularised primal-dual algorithm [6]. The common theme here is that standard interpolation methods are used followed by learned enhancement algorithms.…”

“…We train two networks to map 4 ) . This choice imposes two constraints on the NN design: it must have two input channels on the first layer and a network will always receive the input from the closest acquisition on the first channel of its input layer.…”

Stereo X-Ray Tomography

“…The achievable image quality depends on the number of acquisitions. With the help of advanced algorithms and using regularisation functions such as Total Variation (TV) to enforce certain image smoothness properties [1], or material heterogeneity [2], or learned, datadriven regularisations [3], [4], [5], [6], [7], a significant reduction in the number of measurements is possible, without a significant reduction in image quality. In fact, a range of recent papers has shown that even single projection images are sufficient in certain settings to identify one known image out of a small set of possible reconstructions [8], though this only works by imposing very strong prior constraints, which is only possible if the objects are extremely predictable in their 3D shape.…”