We have developed a novel colorimetric sensor based on a digital camera and white LED illumination. Colorimetric sensor arrays (CSAs) were made from a set of six chemically responsive dyes impregnated on an inert substrate plate by solution casting. Six common amine aqueous solutions, including dimethylamine, triethylamine, diisopropyl-amine, aniline, cyclohexylamine, and pyridine vaporized at 25 °C and six health-related trimethylamine (TMA) concentrations including 170 ppm, 51 ppm, 8 ppm, 2 ppm, 125 ppb and 50 ppb were analyzed by the sensor to test its ability for the qualitative discrimination and quantitative detection of volatile amines. We extracted the feature vectors of the CSA's response to the analytes from a fusional color space, which was obtained by conducting a joint search algorithm of sequential forward selection and sequential backward selection (SFS&SBS) based on the linear discriminant criteria (LDC) in a mixed color space composed of six common color spaces. The principle component analysis (PCA) followed by the hierarchical cluser analysis (HCA) were utilized to discriminate 12 analytes. Results showed that the colorimetric sensor grouped the six amine vapors and five TMA concentrations correctly, while TMA concentrations of 125 ppb and 50 ppb were indiscriminable from each other. The limitation of detection (LOD) of the sensor for TMA was found to be lower than 50 ppb. The CSAs were reusable for TMA concentrations below 8 ppm.

In order to improve the angular resolution of direction-of-arrival (DOA) estimation based on portable microphone at low signal-noise-ratio (SNR), this paper proposes a new high-resolution DOA estimation algorithm based on single acoustic vector sensor (AVS), namely Virtual-AVS algorithm. The innovative algorithm is mainly composed of two parts: the preprocessing method and the DOA estimation method with ambiguity elimination. The preprocessing method is implemented based on higher-order cumulants with Gaussian noise suppression characteristic. This method not only suppress Gaussian noise to effectively enhance SNR, but also achieve AVS virtual expansion. The computational burden can be significantly reduced by effectively expanding one necessary AVS to perform the ESPRIT algorithm. To extend the application range of the innovative algorithm to support non-point-like geometry AVS, we introduce a new DOA estimation method with ambiguity elimination. This method can eliminate the cyclic ambiguity caused by the channel spacing of the non-point-like geometry AVS exceeding the half-wavelength of the incident source. We prove that the Virtual-AVS algorithm is asymptotically unbiased with large number of snapshots. We also derive the Cramér-Rao bound. The simulations of closely-spaced sound sources show that the algorithm has better angular resolution than the existing algorithms. The real-world experiments of semi-anechoic chamber also verify the feasibility of the algorithm.

Existing studies have only qualitatively analyzed the relationship between input power and depth of induction of open-ended coaxial probe. Along with the development of science and technology, the measurement of permittivity requires the quantification of the relationship between the input power and the sensing depth of an open-ended coaxial probe. In this article, for the first time, the effect of different power and materials with different permittivity on the sensing depth of the probe is quantitatively analyzed and the maximum sensing depth of 2.2-mm aperture probe is derived. More importantly, to better handle small amounts of sample materials, thin and multilayer materials, and ensure measurement accuracy, we have also established a sensing depth calculation model. The calculation formula of the model can be used to quickly and accurately calculate the sensing depth of the probe for different permittivity materials under different input powers. This is the industry’s first computational model for the sensing depth of open-end coaxial probes, which provides a strong guarantee of the method’s accurate measurement. Finally, this article validates the model formulation for sensing depth calculation by experimental means. Results show that in the range of −20 to 20 dBm (5.8-GHz frequency) for materials with permittivity between 1 and 80, the computational model is accurately valid and that the maximum depth of sensing for a 2.2-mm aperture probe is 2.8 mm.

Purpose: To propose a synthesis method of pseudo-CT (CT_CycleGAN) images based on an improved 3D cycle generative adversarial network (CycleGAN) to solve the limitations of cone-beam CT (CBCT), which cannot be directly applied to the correction of radiotherapy plans.Methods: The improved U-Net with residual connection and attention gates was used as the generator, and the discriminator was a full convolutional neural network (FCN). The imaging quality of pseudo-CT images is improved by adding a 3D gradient loss function. Fivefold cross-validation was performed to validate our model. Each pseudo CT generated is compared against the real CT image (ground truth CT, CT_gt) of the same patient based on mean absolute error (MAE) and structural similarity index (SSIM). The dice similarity coefficient (DSC) coefficient was used to evaluate the segmentation results of pseudo CT and real CT. 3D CycleGAN performance was compared to 2D CycleGAN based on normalized mutual information (NMI) and peak signal-to-noise ratio (PSNR) metrics between the pseudo-CT and CT_gt images. The dosimetric accuracy of pseudo-CT images was evaluated by gamma analysis.Results: The MAE metric values between the CT_CycleGAN and the real CT in fivefold cross-validation are 52.03 ± 4.26HU, 50.69 ± 5.25HU, 52.48 ± 4.42HU, 51.27 ± 4.56HU, and 51.65 ± 3.97HU, respectively, and the SSIM values are 0.87 ± 0.02, 0.86 ± 0.03, 0.85 ± 0.02, 0.85 ± 0.03, and 0.87 ± 0.03 respectively. The DSC values of the segmentation of bladder, cervix, rectum, and bone between CT_CycleGAN and real CT images are 91.58 ± 0.45, 88.14 ± 1.26, 87.23 ± 2.01, and 92.59 ± 0.33, respectively. Compared with 2D CycleGAN, the 3D CycleGAN based pseudo-CT image is closer to the real image, with NMI values of 0.90 ± 0.01 and PSNR values of 30.70 ± 0.78. The gamma pass rate of the dose distribution between CT_CycleGAN and CT_gt is 97.0% (2%/2 mm).Conclusion: The pseudo-CT images obtained based on the improved 3D CycleGAN have more accurate electronic density and anatomical structure.