Automatic Segmentation and Assessment of Valvular Regurgitations with Color Doppler Echocardiography Images: A VABC-UNet-Based Framework.

This study investigated the automatic segmentation and classification of mitral regurgitation (MR) and tricuspid regurgitation (TR) using a deep learning-based method, aiming to improve the efficiency and accuracy of diagnosis of valvular regurgitations. A VABC-UNet model was proposed consisting of VGG16 encoder, U-Net decoder, batch normalization, attention block and deepened convolution layer based on the U-Net backbone. Then, a VABC-UNet-based assessment framework was established for automatic segmentation, classification, and evaluation of valvular regurgitations. A total of 315 color Doppler echocardiography images of MR and/or TR in an apical four-chamber view were collected, including 35 images in the test dataset and 280 images in the training dataset. In comparison with the classic U-Net and VGG16-UNet models, the segmentation performance of the VABC-UNet model was evaluated via four metrics: Dice, Jaccard, Precision, and Recall. According to the features of regurgitation jet and atrium, the regurgitation could automatically be classified into MR or TR, and evaluated to mild, moderate, moderate-severe, or severe grade by the framework. The results show that the VABC-UNet model has a superior performance in the segmentation of valvular regurgitation jets and atria to the other two models and consequently a higher accuracy of classification and evaluation. There were fewer pseudo- and over-segmentations by the VABC-UNet model and the values of the metrics significantly improved (p < 0.05). The proposed VABC-UNet-based framework achieves automatic segmentation, classification, and evaluation of MR and TR, having potential to assist radiologists in clinical decision making of the regurgitations in valvular heart diseases.

Attention-VGG16-UNet: a novel deep learning approach for automatic segmentation of the median nerve in ultrasound images.

Background: Ultrasonography-an imaging technique that can show the anatomical section of nerves and surrounding tissues-is one of the most effective imaging methods to diagnose nerve diseases. However, segmenting the median nerve in two-dimensional (2D) ultrasound images is challenging due to the tiny and inconspicuous size of the nerve, the low contrast of images, and imaging noise. This study aimed to apply deep learning approaches to improve the accuracy of automatic segmentation of the median nerve in ultrasound images. Methods: In this study, we proposed an improved network called VGG16-UNet, which incorporates a contracting path and an expanding path. The contracting path is the VGG16 model with the 3 fully connected layers removed. The architecture of the expanding path resembles the upsampling path of U-Net. Moreover, attention mechanisms or/and residual modules were added to the U-Net and VGG16-UNet, which sequentially obtained Attention-UNet (A-UNet), Summation-UNet (S-UNet), Attention-Summation-UNet (AS-UNet), Attention-VGG16-UNet (A-VGG16-UNet), Summation-VGG16-UNet (S-VGG16-UNet), and Attention-Summation-VGG16-UNet (AS-VGG16-UNet). Each model was trained on the dataset of 910 median nerve images from 19 participants and tested on 207 frames from a new image sequence. The performance of the models was evaluated by metrics including Dice similarity coefficient (Dice), Jaccard similarity coefficient (Jaccard), Precision, and Recall. Based on the best segmentation results, we reconstructed a 3D median nerve image using the volume rendering method in the Visualization Toolkit (VTK) to assist in clinical nerve diagnosis. Results: The results of paired t-tests showed significant differences (P<0.01) in the metrics' values of different models. It showed that AS-UNet ranked first in U-Net models. The VGG16-UNet and its variants performed better than the corresponding U-Net models. Furthermore, the model's performance with the attention mechanism was superior to that with the residual module either based on U-Net or VGG16-UNet. The A-VGG16-UNet achieved the best performance (Dice =0.904±0.035, Jaccard =0.826±0.057, Precision =0.905±0.061, and Recall =0.909±0.061). Finally, we applied the trained A-VGG16-UNet to segment the median nerve in the image sequence, then reconstructed and visualized the 3D image of the median nerve. Conclusions: This study demonstrates that the attention mechanism and residual module improve deep learning models for segmenting ultrasound images. The proposed VGG16-UNet-based models performed better than U-Net-based models. With segmentation, a 3D median nerve image can be reconstructed and can provide a visual reference for nerve diagnosis.

Study on Different Shear Performance of Moso Bamboo in Four Test Methods.

Bamboo is recognized as a potential and sustainable green material. The longitudinal-splitting and shear strengths of bamboo are weak but critical to its utilizations. To discuss the different shear performances of bamboo, the shear strength and behaviors of bamboo culm were investigated by four test methods: the tensile-shear, step-shear, cross-shear, and short-beam-shear methods. Then, the different shear performance and mechanisms were discussed. Results indicated that the shear strength was significantly different in the four test methods and was highest in the step-shear-test method but lowest in the tensile-shear-test method. Moreover, the typical load-displacement curves were different across the shear methods but were similar to the curves of the respective loading modes. The axially aligned fiber bundles played an important role in all the shear performances. In the tensile-shear method, specimens fractured at the interface of the bamboo-fiber bundles. However, compress-shear behaviors were a combination of compression and shear. Then, the cross-shear method, in compress-shear, was lower than that of the step-shear method because of oval-shaped bamboo culm sections of different thickness. In the short-beam shear method, the behaviors and shearing characteristics were like bending with the fiber bundle pulled out.

Characteristics of malnutrition according to Global Leadership Initiative on Malnutrition criteria in non-surgical patients with inflammatory bowel disease.

OBJECTIVE: Malnutrition is common in patients with inflammatory bowel disease (IBD). The Global Leadership Initiative on Malnutrition (GLIM) was proposed to assess the severity and characteristics of malnutrition. Thus, we aimed to use the latest consensus on the diagnosis of malnutrition, GLIM criteria, to evaluate malnutrition in patients with IBD. METHODS: We performed a retrospective cohort study of 73 adult patients with IBD (48 with Crohn disease and 25 with ulcerative colitis). Demographic data, clinical characteristics, and nutrition status defined by Nutritional Risk Screening (NRS) 2002 and GLIM criteria were recorded at enrollment. RESULTS: According to the GLIM criteria, 43 (58.90%) patients were identified with malnutrition, and the incidence of mild to moderate malnutrition and severe malnutrition was 28.77% (21 of 73 patients) and 30.14% (22 of 73 patients), respectively. The severity of malnutrition in patients with IBD increased with the cumulative number of phenotypic criteria they met (P < 0.01). The difference in the number of etiologic indicators was only identified between patients with severe malnutrition and those without malnutrition (P < 0.05). Patients with Crohns disease had a significantly higher rate of muscle mass loss than patients with ulcerative colitis (P = 0.038) but a lower incidence of reduced food intake or assimilation (P = 0.039). CONCLUSION: The prevalence of malnutrition according to the GLIM criteria was high in non-surgical patients with IBD, and as the degree of malnutrition worsened, more phenotypes and etiologic types appeared. The phenotypic and etiologic characteristics of GLIM were different in patients with Crohn disease than in those with ulcerative colitis.

Properties of Two-Variety Natural Luffa Sponge Columns as Potential Mattress Filling Materials.

Luffa sponge (LS) is a resourceful material with fibro-vascular reticulated structure and extremely high porosity, which make it a potential candidate for manufacturing light mattress. In this study, two types of LS columns, namely high-density (HD) and low-density (LD) columns, were investigated as materials for filling the mattress. The results showed that the compressive strength of HD LS columns was significantly greater than that of LD LS columns. However, the densification strains of the two types of LS column were both in the range of 0.6 to 0.7. Besides, HD LS columns separately pressed to the smooth plateau region and the initial densification region exhibited a partial recovery of instant height when they were unloaded, and then both of them showed no more than 4.2% of height recovery after being allowed to rest at a constant temperature and humidity for 24 h. In contrast, when LD LS columns were compressed to the smooth plateau region, the height recovery was less than 1.62% compared to when they were pressed to the initial densification region, and that was more than 15.62%. Similar to other plant fibers used as mattress fillers, the two types of LS columns also showed good water absorption capacity-both of them could absorb water from as much as 2.07 to 3.45 times their own weight. At the same time, the two types of LS columns also showed good water desorption. The water desorption ratio of HD and LD LS columns separately reached 76.86 and 91.44%, respectively, after being let rest at a constant temperature and humidity for 13 h.