Semi-supervised nuclei segmentation based on multi-edge features fusion attention network.

The morphology of the nuclei represents most of the clinical pathological information, and nuclei segmentation is a vital step in current automated histopathological image analysis. Supervised machine learning-based segmentation models have already achieved outstanding performance with sufficiently precise human annotations. Nevertheless, outlining such labels on numerous nuclei is extremely professional needing and time consuming. Automatic nuclei segmentation with minimal manual interventions is highly needed to promote the effectiveness of clinical pathological researches. Semi-supervised learning greatly reduces the dependence on labeled samples while ensuring sufficient accuracy. In this paper, we propose a Multi-Edge Feature Fusion Attention Network (MEFFA-Net) with three feature inputs including image, pseudo-mask and edge, which enhances its learning ability by considering multiple features. Only a few labeled nuclei boundaries are used to train annotations on the remaining mostly unlabeled data. The MEFFA-Net creates more precise boundary masks for nucleus segmentation based on pseudo-masks, which greatly reduces the dependence on manual labeling. The MEFFA-Block focuses on the nuclei outline and selects features conducive to segment, making full use of the multiple features in segmentation. Experimental results on public multi-organ databases including MoNuSeg, CPM-17 and CoNSeP show that the proposed model has the mean IoU segmentation evaluations of 0.706, 0.751, and 0.722, respectively. The model also achieves better results than some cutting-edge methods while the labeling work is reduced to 1/8 of common supervised strategies. Our method provides a more efficient and accurate basis for nuclei segmentations and further quantifications in pathological researches.

Nuclei segmentation of HE stained histopathological images based on feature global delivery connection network.

The analysis of pathological images, such as cell counting and nuclear morphological measurement, is an essential part in clinical histopathology researches. Due to the diversity of uncertain cell boundaries after staining, automated nuclei segmentation of Hematoxylin-Eosin (HE) stained pathological images remains challenging. Although better performances could be achieved than most of classic image processing methods do, manual labeling is still necessary in a majority of current machine learning based segmentation strategies, which restricts further improvements of efficiency and accuracy. Aiming at the requirements of stable and efficient high-throughput pathological image analysis, an automated Feature Global Delivery Connection Network (FGDC-net) is proposed for nuclei segmentation of HE stained images. Firstly, training sample patches and their corresponding asymmetric labels are automatically generated based on a Full Mixup strategy from RGB to HSV color space. Secondly, in order to add connections between adjacent layers and achieve the purpose of feature selection, FGDC module is designed by removing the jumping connections between codecs commonly used in UNet-based image segmentation networks, which learns the relationships between channels in each layer and pass information selectively. Finally, a dynamic training strategy based on mixed loss is used to increase the generalization capability of the model by flexible epochs. The proposed improvements were verified by the ablation experiments on multiple open databases and own clinical meningioma dataset. Experimental results on multiple datasets showed that FGDC-net could effectively improve the segmentation performances of HE stained pathological images without manual interventions, and provide valuable references for clinical pathological analysis.

Rapid quantitative analysis of potassium in soil based on direct-focused laser ablation-laser induced breakdown spectroscopy.

A fast quantitative analysis method of soil potassium based on direct-focused laser ablation-laser induced breakdown spectroscopy (direct-focused LA-LIBS) was proposed and tested. A high single-pulse energy laser (200 mJ/pulse) beam was focused on the aerosols near the focus of the 10 kHz fiber laser to generate plasma spectra, and the analytical capability of the direct-focused LA-LIBS system was compared with traditional LIBS system using a high single-pulse energy laser (SP-LIBS). The result showed that for moist soil samples the data stability of the direct-focused LA-LIBS method was significantly improved and the R2 factor of the calibration curve improved from 0.64 to 0.93, the limit of detection improved from 159.2 µg/g to 140.9 µg/g. Three random soil samples from different areas of Beijing suburbs were analyzed by the direct-focused LA-LIBS method, and the results were consistent with AAS. The direct-focused LA-LIBS method proposed is different from the traditional double-pulse technology and laser ablation-assisted technology because it not only does not need carrier gas, but also can overcome the matrix differences better, especially the influence of moisture, which provides a new idea for the rapid detection of nutrient elements in field soils.

Epidemiological and molecular characteristics of emergent dengue virus in Yunnan Province near the China-Myanmar-Laos border, 2013-2015.

BACKGROUND: Yunnan Province is located in southwestern China and neighbors the Southeast Asian countries, all of which are dengue-endemic areas. In 2000-2013, sporadic imported cases of dengue fever (DF) were reported almost annually in Yunnan Province. During 2013-2015, we confirmed that a large-scale indigenous DF outbreak emerged in cities of Yunnan Province near the China-Myanmar-Laos border. METHODS: Epidemiological characteristics of DF in Yunnan Province during 2013-2015 were evaluated by retrospective analysis. A total of 232 dengue virus (DENV)-positive sera were randomly collected for sequence analysis of the capsid/premembrane region of DENV from patients with DF in Yunnan Province. The envelope gene of DENV isolates was also amplified and sequenced. Phylogenetic analyses were performed using the neighbor-joining method with the Tajima-Nei model. RESULTS: Phylogenetically, all DENV-positive samples could be classified into DENV-1 genotype I and DENV-2 Asian I genotype during 2013-2015 and DENV-4 genotype I in 2015 from Ruili City; and DENV-3 genotype II in 2013 and DENV-2 Cosmopolitan genotype in 2015 from Xishuangbanna Prefecture. CONCLUSIONS: Our results indicated that imported DF from patients from Laos and Myanmar was the primary cause of the DF epidemic in Yunnan Province. Additionally, DENV strains of all four serotypes were identified in indigenous cases in Yunnan Province during the same time period, while the dengue epidemic pattern observed in southwestern Yunnan showed characteristics of a hypoendemic nature: circulation of DENV-1 and DENV-2 over consecutive years.

[Molecular characteristics of dengue virus outbreak in China-Myanmar border region, Yunnan province, 2015].

OBJECTIVE: To understand the molecular characteristics of a dengue virus outbreak in China-Myanmar border region, Yunnan province, 2015 and provide etiological evidence for the disease control and prevention. METHODS: Semi-nested RTPCR was conducted to detect the capsid premembrane (CprM) gene of RNA of dengue virus by using dengue virus NS1 positive serum samples collected in Mengdin township, Gengma county, Yunnan province in July, 2015. Some positive samples were then detected by using PCR with specific primers to amplify the full E gene. The positive PCR products were directly sequenced. Then sequences generated in this study were BLAST in NCBI website and aligned in Megalign in DNAstar program. Multiple sequence alignments were carried out by using Mega 5.05 software based on the sequences generated in this study and sequences downloaded from GenBank, including the representative strains from different countries and regions. Phylogenetic trees were constructed by using Neighbor-Joining tree methods with Mega 5.05 software. RESULTS: Twenty one of 25 local cases and 10 of 14 imported cases from Myanmar were positive for DENV-1. Eight serum samples were negative for dengue virus. A total of 13 strains with E gene (1485 bp), including 8 local strains and 5 imported strains, were sequenced, which shared 100% nucleotide sequence identities. Twelve strains with CprM gene (406 bp) from 9 local cases and 3 imported cases shared 100% nucleotide sequence identities. Phylogenetic analyses based on E gene showed that the new 13 strains clustered in genotype I of dengue virus and formed a distinct lineage. CONCLUSIONS: This outbreak was caused by genotype I of DENV-1, which had the closest phylogenetic relationships with dengue virus from neighboring Burma area. Comprehensive measures of prevention and control of dengue fever should be strengthened to prevent the spread of dengue virus.