A general deep learning framework for neuron instance segmentation based on Efficient UNet and morphological post-processing.

Recent studies have demonstrated the superiority of deep learning in medical image analysis, especially in cell instance segmentation, a fundamental step for many biological studies. However, the excellent performance of the neural networks requires training on large, unbiased dataset and annotations, which is labor-intensive and expertise-demanding. This paper presents an end-to-end framework to automatically detect and segment NeuN stained neuronal cells on histological images using only point annotations. Unlike traditional nuclei segmentation with point annotation, we propose using point annotation and binary segmentation to synthesize pixel-level annotations. The synthetic masks are used as the ground truth to train the neural network, a U-Net-like architecture with a state-of-the-art network, EfficientNet, as the encoder. Validation results show the superiority of our model compared to other recent methods. In addition, we investigated multiple post-processing schemes and proposed an original strategy to convert the probability map into segmented instances using ultimate erosion and dynamic reconstruction. This approach is easy to configure and outperforms other classical post-processing techniques. This work aims to develop a robust and efficient framework for analyzing neurons using optical microscopic data, which can be used in preclinical biological studies and, more specifically, in the context of neurodegenerative diseases. Code is available at: https://github.com/MIRCen/NeuronInstanceSeg.

High Internal Phase Emulsions Stabilized with Polyphenol-Amyloid Fibril Supramolecules for Encapsulation and Protection of Lutein.

High internal phase emulsions (HIPEs), also called highly concentrated emulsions with a minimal internal phase volume fraction of 74%, have been paid increasing attention in the development of functional foods due to their high potential in loading with large amounts of hydrophobic nutriceuticals. In the present study, HIPEs stabilized by polyphenol-amyloid supramolecular filaments were prepared for encapsulation of olive oil and loading with lutein. Binding and stacking of the green tea polyphenol epigallocatechin gallate (EGCG) on the surface of amyloid fibrils fabricated from hen egg lysozyme resulted in the hybrid supramolecules, which assembled to form hydrogels. The amyloid fibril clusters shrouded by EGCG were observed in the microstructure of the hydrogels characterized by atomic force microscopy (AFM). HIPEs stabilized by the EGCG-amyloid fibril supramolecules showed the typical microstructure of highly packed polyhedral geometric oil droplets. The gel strength of the HIPEs stabilized by the hybrid supramolecules was greater than that of HIPEs stabilized by pure amyloid fibrils. The droplet size of the HIPEs first decreased and then increased with the increase of EGCG contents in the hybrid supramolecules, which was consistent with the corresponding emulsion morphologies obtained from the images of confocal laser scanning microscopy (CLSM). Aggregation of the protein-based nanofibrils appeared in the continuous phase at higher EGCG contents. The droplet size of the HIPEs decreased with the increase of the amyloid fibril concentration, accompanied by more packed and homogenously dispersed lipid droplets, as shown in the CLSM images. A high loading content of lutein of up to 10 mg/mL in the prepared HIPEs was realized, and the stability of lutein against ultraviolet irradiation, heat, iron, and hydrogen peroxide was promoted significantly. In addition, encapsulation with the HIPEs prevented the oxidization of olive oil, and this effect was enhanced with the increase of the EGCG content in the hybrid supramolecules ranging from 0 to 0.25 wt %. The protection function of the HIPEs might be ascribed to the membrane of interfacial amyloid fibrils and the crowded oil droplet environment, both of which could shield the pro-oxidation factors.

Evaluation of Deep Learning Topcoders Method for Neuron Individualization in Histological Macaque Brain Section.

Cell individualization has a vital role in digital pathology image analysis. Deep Learning is considered as an efficient tool for instance segmentation tasks, including cell individualization. However, the precision of the Deep Learning model relies on massive unbiased dataset and manual pixel-level annotations, which is labor intensive. Moreover, most applications of Deep Learning have been developed for processing oncological data. To overcome these challenges, i) we established a pipeline to synthesize pixel-level labels with only point annotations provided; ii) we tested an ensemble Deep Learning algorithm to perform cell individualization on neurological data. Results suggest that the proposed method successfully segments neuronal cells in both object-level and pixel-level, with an average detection accuracy of 0.93.

The feature of cervical microbiota associated with the progression of cervical cancer among reproductive females.

OBJECTIVES: The aim of this study was to characterize cervical microbiome feature of reproductive-age women in the progression of squamous intraepithelial lesions (SIL) to cervical cancer. METHODS: We characterized the 16S rDNA cervical mucus microbiome in 94 participants (age from 18 to 52), including 13 cervical cancer (CA), 31 high-grade SIL (HSIL), 10 low-grade SIL (LSIL), 12 HPV-infected (NH) patients and 28 healthy controls (NN). Alpha (within sample) diversity was examined by Shannon and Simpson index, while Beta (between sample) diversity by principle coordinate analysis (PCoA) of weighted Unifrac distances. Relative abundance of microbial taxa was compared using Linear Discriminant Analysis Effect Size (LEfSe). Co-occurrence analysis was performed to identify correlation among marker genera, and Phylogenetic investigation of communities by reconstruction of unobserved states (PICRUSt) to explore functional features and pathways of cervical microbiota. RESULTS: Alpha diversity(p < 0.05) was higher in severer cervical pathology with lower relative abundance of Lactobacillus as well as higher of anaerobes. Beta diversity (p < 0.01) was significantly different. Marker genera were identified including Porphyromonas, Prevotella and Campylobacter of CA and Sneathia of HSIL. The correlation of differential functional pathways with Prevotella was opposite to that with Lactobacillus. CONCLUSION: Our study suggests differences in cervical microbiota diversity and relative abundance of reproductive-age females in different stages of cervical carcinogenesis. Marker genera might participate in the lesion progression and will be helpful for diagnosis, prevention and treatment. These findings may lead the way to further study of the cervical microbiome in development of cervical cancer.