Search

The ground‐based microwave radiometer (MWR) retrieves atmospheric profiles with a high temporal resolution for temperature and relative humidity up to a height of 10 km. These profiles have been widely used in the field of meteorological observation. Due to the inherent fragility of neural networks, one of the important issues in this field is to improve the reliability and stability of MWR profiles based on neural network inversion. We propose a deep learning method that adds noise to the BP neural network inversion (NBPNN) process. Comparison of the radiosonde data and NBPNN results shows that if the error of MWR brightness temperature is in the range of −2–2 K, the root‐mean‐square error (RMSE) of the temperature profile is 2.15 K, and the RMSE of the relative humidity profile is 19.46 % inverted by NBPNN. The results are much less than the errors of the temperature profile and relative humidity profile inverted by the traditional backpropagation neural network inverse method. From the comparison, we demonstrated that NBPNN significantly increases the inversion accuracy and robustness under the condition of errors in brightness temperature, which can reduce requirements for BT accuracy of MWR and achieve MWR long‐term stability. Key Points: It's necessary to improve the reliability and stability of inversion for microwave radiometerWe propose a deep learning method that adds noise to the BP neural network inversion processThe new inversion process significantly increases the inversion accuracy and robustness under the condition of errors in brightness temperature [ABSTRACT FROM AUTHOR]

Background Esophageal cancer is a prevalent malignancy. There is a considerable demand for developing a fast and noninvasive method to screen out the suspect esophageal cancer patients who may undergo further clinical diagnosis. Methods The exhaled breathes from 29 esophageal cancer patients and 57 healthy people were directly measured using our home-made proton transfer reaction mass spectrometer (PTR-MS). Mann-Whitney U test and stepwise discriminant analysis were applied to identify the ions in the breath mass spectral data which can distinguish cancer cohort from healthy group. Receiver operating characteristics (ROC) analysis was also performed. Results Seven kinds of ions in the breath mass spectrum, viz., m/z 136, m/z 34, m/z 63, m/z 27, m/z 95, m/z 107 and m/z 45, have been found to distinguish between the esophageal cancer patients and healthy people with a sensitivity of 86.2% and a specificity of 89.5%, respectively. Compared with that from the healthy people, the breath mass spectra from esophageal cancer patients show that the mediant intensities of five kinds of ions were decrease and the rest two kinds of ions were increase. ROC analysis gave the area under the curve (AUC) of 0.943. Conclusions This pilot study shows that the ionic characteristics of exhaled VOCs detected by PTR-MS may be used to differentiate between the esophageal cancer patients and the healthy people. Although the breath tests for more patients are needed to confirm such results, the present work indicates that the PTR-MS may be a promising method in the esophageal cancer screening. [ABSTRACT FROM AUTHOR]

Rationale Laser-ablation electrospray ionization mass spectrometry (LAESI-MS) was applied to analyze fresh meat species without sample pretreatment. The study demonstrates that the LAESI-MS technique is a promising, rapid and accurate method for meat identification using a protocol combining principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA). Methods A focused IR-laser was used for meat sample ablation at a wavelength of 2940 nm. The ablated particulates were carried through a transfer PTFE tube using air as carrier gas, delivered to the electrospray plume and ionized. A TOF-MS was used to detect the ion signal. The raw mass spectra were analyzed using the PCA and PLS-DA protocol. Results Five fresh meat samples, chicken, duck, pork, beef and mutton, were identified by the developed LAESI-MS technique using the protocol combining PCA and PLS-DA. The discrimination accuracy of all meat species is 100%, and the score plot also shows good identifying ability. Conclusions Five fresh meat samples were analyzed using the LAESI-MS technique. Each set of raw mass data was collected within 30 s and analyzed by the PCA and PLS-DA protocol. Eighteen, 19, 18, 17, and 15 markers for chicken, duck, pork, beef, and mutton, respectively, have been selected successfully for meat identification. The results demonstrate that LAESI-MS is a new reliable and rapid method for meat identification. Copyright © 2016 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]