Multi-energy CT material decomposition using graph model improved CNN.

In spectral CT imaging, the coefficient image of the basis material obtained by the material decomposition technique can estimate the tissue composition, and its accuracy directly affects the disease diagnosis. Although the precision of material decomposition is increased by employing convolutional neural networks (CNN), extracting the non-local features from the CT image is restricted using the traditional CNN convolution operator. A graph model built by multi-scale non-local self-similar patterns is introduced into multi-material decomposition (MMD). We proposed a novel MMD method based on graph edge-conditioned convolution U-net (GECCU-net) to enhance material image quality. The GECCU-net focuses on developing a multi-scale encoder. At the network coding stage, three paths are applied to capture comprehensive image features. The local and non-local feature aggregation (LNFA) blocks are designed to integrate the local and non-local features from different paths. The graph edge-conditioned convolution based on non-Euclidean space excavates the non-local features. A hybrid loss function is defined to accommodate multi-scale input images and avoid over-smoothing of results. The proposed network is compared quantitatively with base CNN models on the simulated and real datasets. The material images generated by GECCU-net have less noise and artifacts while retaining more information on tissue. The Structural SIMilarity (SSIM) of the obtained abdomen and chest water maps reaches 0.9976 and 0.9990, respectively, and the RMSE reduces to 0.1218 and 0.4903 g/cm3. The proposed method can improve MMD performance and has potential applications.

A novel fluorescent probe of alkyne compounds for putrescine detection based on click reaction.

Putrescine is a toxic biogenic amine produced in the process of food spoilage, and a high concentration of biogenic amines in foods will cause health problems such as abnormal blood pressure, headaches and tachycardia asthma/worsening asthma. The detection of putrescine is necessary. However, traditional putrescine detection requires specialized instruments and complex operations. To detect putrescine quickly, sensitively and accurately, we designed and successfully prepared a fluorescent probe (DPY) with active alkynyl groups. DPY takes p-dimethoxybenzene as the raw material, adding a highly active alkyne group. It is stable in experimental pH (∼7) because the UV-vis absorption and fluorescence emission spectra in pH = 3-12 have little change. The fluorescence intensity of DPY decreased only about 1% under the irradiation of 420 nm within 2 h, showing its better photostability. DPY has a high selectivity to putrescine because of the amino-alkyne click reaction without any catalyst in presence of different biogenic amines. The obvious response to putrescine was found in 30 seconds at room temperature. The mechanism between DPY and putrescine was investigated before and after adding putrescine by 1H NMR spectra and the Job plot. The results indicated a typical 1 : 1 stoichiometry between the DPY and DAB. Furthermore, the higher sensitivity of DPY to putrescine was obtained with the detection of limit (LOD) of 3.19 × 10-7 mol L-1, which was better than that of the national standard (2.27 × 10-5 mol L-1). The novel fluorescent probe was successfully applied to beer samples to detect putrescine. The proposed strategy is expected to provide some guidance for the development of some new ways to detect food security.

A semi-supervised learning method of latent features based on convolutional neural networks for CT metal artifact reduction.

PURPOSE: X-ray computed tomography (CT) has become a convenient and efficient clinical medical technique. However, in the presence of metal implants, CT images may be corrupted by metal artifacts. The metal artifact reduction (MAR) methods based on deep learning are mostly supervised methods trained with labeled synthetic-artifact CT images. However, this causes the neural network to be biased toward learning specific synthetic-artifact patterns and leads to a poor generalization for unlabeled real-artifact CT images. In this study, a semi-supervised learning method of latent features based on convolutional neural networks (SLF-CNN) is developed to remove metal artifacts while ensuring a good generalization ability for real-artifact CT images. METHODS: The proposed semi-supervised method extracts CT image features in alternate iterations of a synthetic-artifact learning stage and a real-artifact learning stage. In the synthetic-artifact learning stage, SLF-CNN is fed with paired synthetic-artifact CT images and is constrained using mean-squared-error (MSE) loss and perceptual loss in a supervised learning fashion. In the real-artifact learning stage, the network weight is updated by minimizing the error between the pseudo-ground truths and the predicted latent features. The feature level pseudo-ground truths are obtained by modeling latent features using the Gaussian process. The overall framework of SLF-CNN adopts an encoder-decoder structure. The encoder is composed of artifact information collection groups to map the input artifact-affected synthetic-artifact CT images and real-artifact CT images into latent features. The decoder is composed of stacked ResNeXt blocks and is responsible for decoding latent features with high-level semantic information to reconstruct artifact-free CT images. The performance of the proposed method is evaluated through contrast experiments and ablation experiments. RESULTS: The contrast experimental results indicate that the artifact-free CT images obtained by SLF-CNN have good metrics values, which are close to or better than those of typical supervised MAR methods. The metal artifacts in artifact-affected CT images are eliminated and the tissue structure details are preserved using SLF-CNN. The ablation experiment shows that adding real-artifact CT images greatly improves the generalization ability of the network. CONCLUSIONS: The proposed semi-supervised learning method of latent features for MAR effectively suppresses metal artifacts and improves the generalization ability of the network.

An x-ray crosstalk correction method using FCNN for a novel energy resolving scheme in spectral CT.

Spectral computed tomography has great potential for multi-energy imaging and anti-artifacts. The complete absorption-based energy resolving scheme of x-rays has been used for the integrity of detected information. However, this scheme is limited by the fact that the detector pixel thickness is high and fixed. Here, an energy resolving scheme is proposed using the crosstalk correction method for the incomplete absorption detection of x-rays. A fully connected neural network (FCNN)-based method was used to correct the difference caused by internal x-ray crosstalk of the edge-on detector. The energy and spatial features of the data which is collected in layers were combined to establish the mapping between the ideal data and the data with crosstalk at the pre-processing stage. Thereafter, to reconstruct the stable and highly accurate energy-resolving equations, the layers with low relative energy difference were selected and grouped together to reduce the accumulation difference. The experiment results demonstrate the feasibility of this energy resolving scheme. The differences caused by crosstalk can be suppressed through the proposed FCNN-based method. The resolving accuracy can be further improved by grouping more layers at forward positions in the pixel. Moreover, this improvement can be observed in the reconstructed images with reduced artifacts and improved quality.