Compressive x-ray tomosynthesis (CXT) uses a set of encoded projection measurements from different incident angles to reconstruct the object under inspection. We consider the variable motion of objects on a conveyor mechanism and establish an imaging model based on the sensing geometry of a dynamic CXT system. Then, a numerical algorithm is proposed to optimize the structured illumination series to improve reconstruction accuracy with reduced radiation dose. Compared with the state-of-the-art method, the proposed strategy increases the degrees of optimization freedom by jointly optimizing the coding mask patterns, locations of x-ray sources, and exposure moments in the CXT system, thus obtaining better reconstruction performance. A genetic algorithm is applied to achieve the optimization results. It shows that the proposed method outperforms the traditional CXT approach by further improving reconstruction performance under comparable radiation dose.
Dynamic coded x-ray tomosynthesis (CXT) uses a set of encoded x-ray sources to interrogate objects lying on a moving conveyor mechanism. The object is reconstructed from the encoded measurements received by the uniform linear array detectors. We propose a multi-objective optimization (MO) method for structured illuminations to balance the reconstruction quality and radiation dose in a dynamic CXT system. The MO framework is established based on a dynamic sensing geometry with binary coding masks. The Strength Pareto Evolutionary Algorithm 2 is used to solve the MO problem by jointly optimizing the coding masks, locations of x-ray sources, and exposure moments. Computational experiments are implemented to assess the proposed MO method. They show that the proposed strategy can obtain a set of Pareto optimal solutions with different levels of radiation dose and better reconstruction quality than the initial setting.
more » « less- PAR ID:
- 10275808
- Publisher / Repository:
- Optical Society of America
- Date Published:
- Journal Name:
- Applied Optics
- Volume:
- 60
- Issue:
- 21
- ISSN:
- 1559-128X; APOPAI
- Format(s):
- Medium: X Size: Article No. 6177
- Size(s):
- Article No. 6177
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
-
more » « less
Coded x-ray diffraction imaging (CXRDI) is an emerging computational imaging approach that aims to solve the phase retrieval problem in x-ray crystallography based on the intensity measurements of encoded diffraction patterns. Boolean coding masks (BCMs) with complementary structures have been used to modulate the diffraction pattern in CXRDI. However, the optimal spatial distribution of BCMs still remains an open problem to be studied in depth. Based on the spectral initialization criterion, we provide a theoretical proof for the premise that the optimal complementary BCMs should obey the blue noise distribution in the sense of mathematical expectation. In addition, the benefits of the blue noise coding strategy are assessed by a set of simulations, where better reconstruction quality is observed compared to the random BCMs and other complementary BCMs.
-
more » « less
Compressive spectral X-ray imaging (CSXI) introduces a pixelated spectral modulator called K-edge coded aperture (KCA) in front of the X-ray source, which enables both, lower dosage to the subject, as well as the capability of spectral tomography while using low-cost integrating X-ray detectors. CSXI systems generally use hundreds of different spectral modulators, each with a distinct pattern to uniquely modulate the illumination at every view angle. In contrast, this paper introduces the use of a single and static coded aperture placed in a tomosynthesis gantry. The compressive system thus interrogates the subject with a fixed coded illumination pattern on all view angles. The advantages of the system are many including reduced cost and the feasibility of implementation. Given the reduced set of coded measurement and the limited spectral separation ability in the resulting architecture, the nonlinear inverse reconstruction problem results in a highly ill-posed problem. An efficient alternating minimization method with three-dimensional total variation regularization is developed for image reconstruction. Furthermore, rather than simply using a random pattern, the coded aperture is optimized under a uniform sensing criterion that shapes the spatial and spectral pattern of the coded aperture so as to minimize the overall radiation exposure placed on any volumetric area of the patient. This is of particular importance in medical imaging where patients at risk are recommended to have periodical X-ray tomosynthesis screenings. The coded aperture optimization is then posed as a binary programming problem solved by a gradient-based algorithm with equilibrium constraints. Numerical experiments show that spatial and spectral coding used in the proposed system to interrogate the subject not only reduces the radiation dose but it also improves the quality of image reconstruction. Gains close to 5dB in peak signal to noise ratio are observed in simulations. Furthermore, it is shown that the optimization of the KCA can effectively improve the uniformity of X-ray radiation compared to random KCA modulation, thus reducing the radiation dose throughout all volumetric sub-areas of the subject — an objective that is not possible with the use of random KCAs.
-
more » « less
Compressive X-ray tomosynthesis uses a few two-dimensional projection measurements modulated by coding masks to reconstruct the three-dimensional object that can be sparsely represented on a predefined basis. However, the coding mask optimization and object reconstruction require significant computing resources. In addition, existing methods fall short to exploits the synergy between the encoding and reconstruction stages to approach the global optimum. This paper proposes a model-driven deep learning (MDL) approach to significantly improve the computational efficiency and accuracy of tomosynthesis reconstruction. A unified framework is developed to jointly optimize the coding masks and the neural network parameters, which effectively increase the degrees of optimization freedom. It shows that the computational efficiency of coding mask optimization and image reconstruction can be improved by more than one order of magnitude. Furthermore, the performance of reconstruction results is significantly improved.
-
more » « less
Coded aperture X-ray CT (CAXCT) is a new low-dose imaging technology that promises far-reaching benefits in industrial and clinical applications. It places various coded apertures (CA) at a time in front of the X-ray source to partially block the radiation. The ill-posed inverse reconstruction problem is then solved using l1-norm-based iterative reconstruction methods. Unfortunately, to attain high-quality reconstructions, the CA patterns must change in concert with the view-angles making the implementation impractical. This paper proposes a simple yet radically different approach to CAXCT, which is coined StaticCodeCT, that uses a single-static CA in the CT gantry, thus making the imaging system amenable for practical implementations. Rather than using conventional compressed sensing algorithms for recovery, we introduce a new reconstruction framework for StaticCodeCT. Namely, we synthesize the missing measurements using low-rank tensor completion principles that exploit the multi-dimensional data correlation and low-rank nature of a 3-way tensor formed by stacking the 2D coded CT projections. Then, we use the FDK algorithm to recover the 3D object. Computational experiments using experimental projection measurements exhibit up to 10% gains in the normalized root mean square distance of the reconstruction using the proposed method compared with those attained by alternative low-dose systems.
