KEYWORDS: Image processing, Data processing, Denoising, Relative intensity noise, Data transmission, X-rays, Signal to noise ratio, Signal attenuation, Medical imaging
Photon counting detector (PCD) is a hot topic at present. Compared with traditional energy integral detector, it has the potential of high spatial resolution, high sensitivity and low dose, which can effectively promote medical imaging diagnosis. However, when PCD is counting X-ray photons, the photon number of each energy bin is relatively small. Additionally, charge-sharing response and pulse superposition effect will also affect the photon count rate, resulting in serious noise and affecting the imaging quality. In this paper, a photon-counting denoising algorithm based on subspace decomposition is proposed. According to the similarity between the data of different bins and the self-similarity of the data, this paper constructs sparse representation by subspace decomposition method and uses block matching algorithm to suppress noise. In simulation experiments, we carried out spectral computed tomography imaging experiments with the three-dimensional phantom of a digital mice based on PCD, and denoised the data by different algorithms. The quantitative results show that our method improves peak signal-to-noise ratio by 2.21dB compared with block-matching and 3D filtering when photon flux is 4×103 , which verifies the potential of the proposed algorithm in medical imaging.
Spectral computed tomography (spectral CT) is an emerging imaging technology that is capable of distinguishing material properties. However, the difficulty of decomposition process is intensified by the nonlinearity of the measurements and ill-conditioned problem, particularly when the number of materials and energies do not match. Therefore, the key issue in spectral CT is to design an improved algorithm in the accurate material decomposition. This study proposes a one-step multi-material algorithm that combines a statistical reconstruction model with a gradient-sparsity based prior. In our approach, the elimination based method under volume conservation constrained condition is designed for the inconsistent scanning of number of materials and energies. Newton descent method is adopted to efficiently solve the optimization problem based on a simple surrogate function. Through simulated experiments, the proposed method achieves a significantly higher peak signal-to-noise ratio (PSNR) compared to other algorithms, with an increase of approximately 23.988 dB and 23.462 dB. Numerical experiments have confirmed the efficiency of our proposed method in reconstructing the material distributions while reducing noise compared to state-of-the-art methods.
Limited-angle computed tomography (CT) image reconstruction is a challenging reconstruction problem in the fields of CT. With the development of deep learning, the generative adversarial network (GAN) perform well in image restoration by approximating the distribution of training sample data. In this paper, we proposed an effective GAN-based inpainting method to restore the missing sinogram data for limited-angle scanning. To estimate the missing data, we design the generator and discriminator of the patch-GAN and train the network to learn the data distribution of the sinogram. We obtain the reconstructed image from the restored sinogram by filtered back projection and simultaneous algebraic reconstruction technique with total variation. Experimental results show that serious artifacts caused by missing projection data can be reduced by the proposed method, and it is hopeful to solve the reconstruction problem of 60° limited scanning angle.
The improvement of computed tomography (CT) image resolution is beneficial to the subsequent medical diagnosis, but it is usually limited by the scanning devices and great expense. Convolutional neural network (CNN)- based methods have achieved promising ability in super-resolution. However, existing methods mainly focus on the super-resolution of reconstructed image and do not fully explored the approach of super-resolution from projectiondomain. In this paper, we studied the characteristic of projection and proposed a CNN-based super-resolution method to establish the mapping relationship of low- and high-resolution projection. The network label is high-resolution projection and the input is its corresponding interpolation data after down sampling. FDK algorithm is utilized for three-dimensional image reconstruction and one slice of reconstruction image is taken as an example to evaluate the performance of the proposed method. Qualitative and quantitative results show that the proposed method is potential to improve the resolution of projection and enables the reconstructed image with higher quality.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.