HGR-Net: Hierarchical Graph Reasoning Network for Arbitrary Shape Scene Text Detection.

As a prerequisite step of scene text reading, scene text detection is known as a challenging task due to natural scene text diversity and variability. Most existing methods either adopt bottom-up sub-text component extraction or focus on top-down text contour regression. From a hybrid perspective, we explore hierarchical text instance-level and component-level representation for arbitrarily-shaped scene text detection. In this work, we propose a novel Hierarchical Graph Reasoning Network (HGR-Net), which consists of a Text Feature Extraction Network (TFEN) and a Text Relation Learner Network (TRLN). TFEN adaptively learns multi-grained text candidates based on shared convolutional feature maps, including instance-level text contours and component-level quadrangles. In TRLN, an inter-text graph is constructed to explore global contextual information with position-awareness between text instances, and an intra-text graph is designed to estimate geometric attributes for establishing component-level linkages. Next, we bridge the cross-feed interaction between instance-level and component-level, and it further achieves hierarchical relational reasoning by learning complementary graph embeddings across levels. Experiments conducted on three publicly available benchmarks SCUT-CTW1500, Total-Text, and ICDAR15 have demonstrated that HGR-Net achieves state-of-the-art performance on arbitrary orientation and arbitrary shape scene text detection.

Prediction of microvascular invasion in hepatocellular carcinoma with magnetic resonance imaging using models combining deep attention mechanism with clinical features / 南方医科大学学报

OBJECTIVE@#To investigate the consistency and diagnostic performance of magnetic resonance imaging (MRI) for detecting microvascular invasion (MVI) of hepatocellular carcinoma (HCC) and the validity of deep learning attention mechanisms and clinical features for MVI grade prediction.@*METHODS@#This retrospective study was conducted among 158 patients with HCC treated in Shunde Hospital Affiliated to Southern Medical University between January, 2017 and February, 2020. The imaging data and clinical data of the patients were collected to establish single sequence deep learning models and fusion models based on the EfficientNetB0 and attention modules. The imaging data included conventional MRI sequences (T1WI, T2WI, and DWI), enhanced MRI sequences (AP, PP, EP, and HBP) and synthesized MRI sequences (T1mapping-pre and T1mapping-20 min), and the high-risk areas of MVI were visualized using deep learning visualization techniques.@*RESULTS@#The fusion model based on T1mapping-20min sequence and clinical features outperformed other fusion models with an accuracy of 0.8376, a sensitivity of 0.8378, a specificity of 0.8702, and an AUC of 0.8501 for detecting MVI. The deep fusion models were also capable of displaying the high-risk areas of MVI.@*CONCLUSION@#The fusion models based on multiple MRI sequences can effectively detect MVI in patients with HCC, demonstrating the validity of deep learning algorithm that combines attention mechanism and clinical features for MVI grade prediction.

[Cascaded multi-level medical image registration method based on transformer].

In deep learning-based image registration, the deformable region with complex anatomical structures is an important factor affecting the accuracy of network registration. However, it is difficult for existing methods to pay attention to complex anatomical regions of images. At the same time, the receptive field of the convolutional neural network is limited by the size of its convolution kernel, and it is difficult to learn the relationship between the voxels with far spatial location, making it difficult to deal with the large region deformation problem. Aiming at the above two problems, this paper proposes a cascaded multi-level registration network model based on transformer, and equipped it with a difficult deformable region perceptron based on mean square error. The difficult deformation perceptron uses sliding window and floating window techniques to retrieve the registered images, obtain the difficult deformation coefficient of each voxel, and identify the regions with the worst registration effect. In this study, the cascaded multi-level registration network model adopts the difficult deformation perceptron for hierarchical connection, and the self-attention mechanism is used to extract global features in the basic registration network to optimize the registration results of different scales. The experimental results show that the method proposed in this paper can perform progressive registration of complex deformation regions, thereby optimizing the registration results of brain medical images, which has a good auxiliary effect on the clinical diagnosis of doctors.

Present status and future prospect of accreditation of proficiency testing provider in the area of laboratory medicine in China / 中华检验医学杂志

The accreditation of Proficiency Testing Provider ( PTP) has gone through more than a decade in China.Over the past ten years, from nonexistence to existence, the domestic PTPs in the area of laboratory medicine have gradually standardized and developed.By reviewing the international and domestic practice, this paper gives an outline of the present status of the accreditation of PTP in the area of laboratory medicine, explores the difference with other countries, the main problems and some suggestions for improvement, and makes a prospect of the development of the accreditation of PTP in the area of laboratory medicine in China.

Protective Role of Rheum Tanguticum Polysaccharide 1 in Radiation- induced Intestinal Mucosal Injury.

The protective effects of Rheum tanguticum polysaccharide 1 (RTP1), which is extracted from the Chinese traditional medicine Rheum tanguticum, on radiation-induced intestinal mucosal injury was investigated. Rat intestinal crypt epithelial cells (IEC-6 cells) and Sprague-Dawley rats were each divided into control, irradiated and RTP1-pretreated irradiated groups. After irradiation, cell survival was determined by MTT (3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide). assay, and the intracellular reactive oxygen species (ROS) was detected by fluorescent probe method. Apoptosis was observed by acridine orange staining, and cell cycle was analysed by flow cytometry. Histological analysis of the rat intestinal mucosa was conducted by haematoxylin and eosin staining. Irradiation at 8 Gy(Gray) decreased cell survival rate to only 54%, significantly increased intracellular ROS levels and induced apoptosis. RTP1 pretreatment significantly inhibited cell death, reduced the formation of intracellular ROS and partially inhibited apoptosis. Irradiation markedly reduced the height and quantity of rat intestinal villi, but it could be antagonised by RTP1 pretreatment. RTP1 can promote the recovery of intestinal mucosa damage, possibly by inhibiting radiation-induced intestinal epithelial apoptosis and intracellular ROS production.

Regulation of RAGE splicing by hnRNP A1 and Tra2ß-1 and its potential role in AD pathogenesis.

The receptor for advanced glycation end products (RAGE) gene expresses two major alternative splicing isoforms, full-length membrane-bound RAGE (mRAGE) and secretory RAGE (esRAGE). Both isoforms play important roles in Alzheimer's disease (AD) pathogenesis, either via interaction of mRAGE with ß-amyloid peptide (Aß) or inhibition of the mRAGE-activated signaling pathway. In the present study, we showed that heterogeneous nuclear ribonucleoprotein A1 (hnRNP A1) and Transformer2ß-1 (Tra2ß-1) were involved in the alternative splicing of mRAGE and esRAGE. Functionally, two factors had an antagonistic effect on the regulation. Glucose deprivation induced an increased ratio of mRAGE/esRAGE via up-regulation of hnRNP A1 and down-regulation of Tra2ß-1. Moreover, the ratios of mRAGE/esRAGE and hnRNP A1/Tra2ß-1 were increased in peripheral blood mononuclear cells from AD patients. The results provide a molecular basis for altered splicing of mRAGE and esRAGE in AD pathogenesis. The receptor for advanced glycation end products (RAGE) gene expresses two major alternative splicing isoforms, membrane-bound RAGE (mRAGE) and secretory RAGE (esRAGE). Both isoforms play important roles in Alzheimer's disease (AD) pathogenesis. Mechanism for imbalanced expression of these two isoforms in AD brain remains elusive. We proposed here a hypothetic model to illustrate that impaired glucose metabolism in AD brain may increase the expression of splicing protein hnRNP A1 and reduce Tra2ß-1, which cause the imbalanced expression of mRAGE and esRAGE.

Determination of urine oxalate level in rats with renal calcium oxalate calculus by high-performance liquid chromatography / 中国医学科学院学报

<p><b>OBJECTIVE</b>To establish a method of high-performance liquid chromatography (HPLC) for determining the urine oxalate levle in rats with renal calcium oxalate calculus.</p><p><b>METHODS</b>Totally 24 SPF Wistar healthy male rats were randomly divided into control group(n=12)and ethylene glycol (EG) group (n=12). Rats in EG group were administered intragastrically with 2% ammonium chloride (AC)2 ml/rat per day+1% ethylene glycol (EG), along with free access to drinking water.The control group was fed with deionized water, along with the intragastric administration of normal saline (1 ml per day). Twenty-eight days after modelling, the 24-hour urine samples were collected, and the urine oxalic acid levels were determined using HPLC and the results were compared with those of catalytic spectrophotometry using oxidation of methyl. During the HPLC, the samples were separated on Aglient 5TC-C18 (250×4.6 mm,5 Μm), eluted with mixture of methanol (0.1 mol/L) and ammonium acetate (15:85) at 1.2 ml/min, and detected at 314 nm, with the column temperature being 20 ℃.</p><p><b>RESULTS</b>The standard curves of high and low concentrations of oxalic acid were y=5909.1x+378730, R² =0.9984 and y=7810.5x-16635, R² =0.9967,respectively. The lowest detectable concentration in this method was 5 Μg/ml. The linear high concentration range of oxalate stood at 62.50-2000.00 Μg/ml, and the linear low concentration range of oxalate stood at 6.25-100.00 Μg/ml. Its average recovery was 95.1%, and its within-day and day-to-day precisions were 3.4%-10.8% and 3.8%-9.4%. Both HPLC and catalytic spectrophotometry showed significantly higher urinary oxalic acid concentration and 24 h urine oxalate level in EG group compared with the control group [urinary oxalic acid concentration: (736.35 ± 254.52) Μg/ml vs.(51.56 ± 36.34) Μg/ml,(687.35 ± 234.53) Μg/ml vs.(50.24 ± 42.34) Μg/ml;24 h urine oxalate level: (11.23 ± 4.12)mg vs.(0.87 ± 0.45)mg,(9.89 ± 3.55)mg vs. (0.77 ± 0.65)mg; all P<0.01]. No statistically significant difference was observed in the results of urinary oxalate concentration and 24 h urine oxalate level between HPLC and potassium chromate oxidation of methyl red spectrophotometry (all P>0.05).</p><p><b>CONCLUSION</b>HPLC is a simple, rapid, and precise method in detecting urine oxalate level in rats with renal calcium oxalate calculus, with high recovery rate.</p>