Search

Sensitive skin is described as an unpleasant sensory response to a stimulus that should not cause a sensation. Sensitive skin affects an increasing proportion of the population. Sixty-seven participants who tested positive to lactic acid sting test were recruited and randomized into two groups to observe the clinical efficacy and safety of a new birch juice spray for repairing sensitive skin. One group used test spray A, while the other group used spray B as a control. Both groups were sprayed six times daily for 28 days. Noninvasive testing instruments were used to measure stratum corneum hydration, sebum content, transepidermal water loss rates, skin blood perfusion and current perception threshold before and after using spray. Facial images were captured by VISIA-CR, and the image analysis program (Image‐Pro Plus) was used to analyze these to obtain the redness value of the facial skin. Moreover, lactic acid sting test scores and participants' self-assessments were also performed at baseline, week 2 and week 4. Both sprays A and B significantly decreased the lactic acid sting test score, transepidermal water loss rates, skin blood perfusion, and redness, while increasing the stratum corneum hydration. Compared to spray B, spray A increased sensory nerve thresholds at 5 Hz and decreased the transepidermal water loss rates, skin blood perfusion, and lactic acid sting test score. Sprays containing birch juice improved cutaneous biophysical properties in participants with sensitive skin. [ABSTRACT FROM AUTHOR]

Objective: With the constant upgrading of healthcare concepts, silicone‐free hair products have attracted more attention among consumers. In the present study, transparent silicone‐free shampoo containing proteins was successfully fabricated by mixing mild non‐sulphate surfactants, compound cationic conditioners, natural silicone oil substitutes, protein conditioners, thickeners, and other auxiliary ingredients. The effects of the type of surfactants, hair conditioners and thickeners, the type and content of proteins, and the mass ratio of compound proteins on the hair foaming performance, hair grooming performance, and penetration performance were investigated. Methods: The basic formulation framework for transparent silicone‐free shampoo was established at first. Then, various hydrolyzed proteins were further added to the basic formulation in the form of single use or compound use to prepare transparent silicone‐free shampoo containing different proteins. The morphology of hair samples and penetration of protein in hair were evaluated with a scanning electron microscope and laser confocal fluorescence microscopy, respectively. And the hair grooming performance was also determined by a dynamic combing tester. Results: The compound proteins of Croquat WKP PE‐LQ‐WD and Gluadin® Kera‐PLM with mass ratios of 1:9 and 1:1 at 2 wt% total protein content added to the silicone‐free shampoo brought a higher contact angle and a lower frictional coefficient than commercial silicone‐free shampoo without proteins. In addition, the compound proteins also have the dual effect of adsorbing the hair surface and penetrating deep into the interior of the hair. Conclusion: The combination of cationic modified hydrolyzed protein (e.g., Croquat WKP PE‐LQ‐WD) and hydrolyzed protein with low molecular weight (e.g., Gluadin® Kera‐P LM) at an appropriate mass ratio exhibited a strong synergistic effect on hair conditioning properties. It could provide a significant reference for developing silicone‐free hair products with more benefits. [ABSTRACT FROM AUTHOR]

High-precision numerical simulation for solving unsteady flow fields is both time- and labor-intensive. Additionally, it is challenging to directly apply to multidisciplinary design fields, such as multi-variable optimization and stability prediction. In this study, a multi-scale mixed information-driven hybrid deep neural network (HDNN) model is proposed to realize fast prediction of unsteady flow fields. The proposed model contains two modules: (a) an HDNN comprising convolutional, convolutional gated recurrent unit, and deconvolutional layers for the modeling of the spatio-temporal dynamics of the flow fields; and (b) a multi-scale mixed loss (MSML) function to evaluate the localized prediction error and feature similarity between the predicted and actual flow fields at different spatial scales. We constructed two datasets, one comprising the flow around a two-dimensional cylinder and the other comprising the flow past an airfoil, respectively. These datasets were utilized for model validation. Experimental results demonstrated that the model predictions aligned well with the computational fluid dynamics simulation results. Furthermore, the computation time of the model is only 10.5% of that required for CFD. We further demonstrated the efficacy of the HDNN model and MSML function through ablation experiments. • A novel model for rapid prediction of unsteady flow fields is proposed. • A loss function called multi-scale mixed loss is proposed. • The proposed model efficiently predicts unsteady flows with reasonable accuracy. • The proposed model can continuously predict unsteady flows based on its prediction. [ABSTRACT FROM AUTHOR]

In recent years, deep learning has dramatically improved the cognitive ability of the network by extracting depth features, and has been successfully applied in the field of feature extraction and classification of hyperspectral images. However, it is facing great difficulties for target detection due to extremely limited available labeled samples that are insufficient to train deep networks. In this paper, a novel target detection framework for deep learning is proposed, denoted as HTD-Net. To overcome the few-training-sample issue, the proposed framework utilizes an improved autoencoder (AE) to generate target signatures, and then finds background samples which differ significantly from target samples based on a linear prediction (LP) strategy. Then, the obtained target and background samples are used to enlarge the training set by generating pixel-pairs, which is viewed as the input of a pre-designed network architecture to learn discriminative similarity. During testing, pixel-pairs of a pixel to be labeled are constructed with both available target samples and background samples. Spectral difference between these pixel-pairs is classified by the well-trained network with results of similarity measurement. The outputs from a two-branch averaged similarity scores are combined to generate the final detection. Experimental results with several real hyperspectral data demonstrate the superiority of the proposed algorithm compared to some traditional target detectors. [ABSTRACT FROM AUTHOR]