Search

Many major diseases can cause changes in the morphology of blood vessels, and the segmentation of retinal blood vessels is of great significance for preventing these diseases. Obtaining complete, continuous, and high-resolution segmentation results is very challenging due to the diverse structures of retinal tissues, the complex spatial structures of blood vessels, and the presence of many small ships. In recent years, deep learning networks like UNet have been widely used in medical image processing. However, the continuous down-sampling operations in UNet can result in the loss of a significant amount of information. Although skip connections between the encoder and decoder can help address this issue, the encoder features still contain a large amount of irrelevant information that cannot be efficiently utilized by the decoder. To alleviate the irrelevant information, this paper proposes a feature selection module between the decoder and encoder that utilizes the self-attention mechanism of transformers to accurately and efficiently select the relevant encoder features for the decoder. Additionally, a lightweight Residual Global Context module is proposed to obtain dense global contextual information and establish dependencies between pixels, which can effectively preserve vascular details and segment small vessels accurately and continuously. Experimental results on three publicly available color fundus image datasets (DRIVE, CHASE, and STARE) demonstrate that the proposed algorithm outperforms existing methods in terms of both performance metrics and visual quality. [ABSTRACT FROM AUTHOR]

Shale oil and gas are currently the major fields of unconventional hydrocarbon exploration and development. The Fengcheng Formation (FF) shale in the Mahu Sag of the Junggar Basin is an alkaline lacustrine organic-rich shale with an extremely prospective shale oil potential. However, its strong heterogeneity and complex pore structure greatly influence the development of shale oil. It is significant to investigate the pore and geochemical characteristics of shale reservoirs for shale oil extraction. In this study, the pore structure and geochemical characteristics of FF have been investigated using core analysis, Rock-Eval pyrolysis, X-ray diffraction (XRD), field-emission scanning electron microscopy (FE-SEM), mercury injection capillary pressure (MICP), low-temperature gas adsorption (LTGA), and X-ray computed tomography (X-ray CT). The results show that the shale of FF has moderate organic matter abundance, and the kerogen is mainly of Type II, which is now at the peak of oil generation. Shale minerals are mainly composed of carbonate (dolomite and calcite) and siliceous (quartz and feldspar) minerals, with extremely low clay mineral content. The pore types are mainly intergranular pores (inter-P), intragranular pores (intra-P), and microfractures associated with mineral particles. The pore space is contributed predominantly by micropores of 0.5–1.2 nm and mesopores of 10–50 nm, whereas macropores are underdeveloped. The pores are mostly ink bottle- and slit-shaped, and the pore connectivity is relatively poor. The pore development of shale in the FF is influenced by organic matter abundance, thermal maturity, mineral composition, etc. Organic matter content (TOC), thermal maturity (Ro), and carbonate minerals have a positive effect on pore development, and the pore volume (PV) increases with TOC, Ro, and carbonate minerals. While clay minerals show a negative effect, the PV decreases with clay minerals. Additionally, the influence of the clay mineral content on the pore morphology of shale should not be ignored. This study investigates the pore structure and geochemical characteristics of the alkaline lacustrine shale of FF in Mahu Sag, which is significant to deepen the understanding of alkaline lacustrine shale and to improve the production of shale oil. [ABSTRACT FROM AUTHOR]

Uncooled focal plane arrays (FPAs) are being developed for a wide range of infrared imaging applications. A substrate-free FPA for optical readout infrared imaging is fabricated with a pixel pitch of 120 μm. The pressure dependences of thermal conductance of a FPA with/without substrate are studied by modeling analysis. Infrared imaging experiments are performed to validate the modeling analysis. At atmospheric pressure the total thermal conductance of the substrate-free FPA is only 1/1000 time of the traditional FPA which has a 2 μm air gap between the cantilever beam and the substrate. Room temperature object’s thermal images are obtained even if the FPA is placed in the atmosphere. The air conductance of a substrate-free FPA at atmospheric pressure could be as small as that of a traditional FPA with a substrate in high vacuum (about 1 Pa). The experimental result also shows that the system noise keeps almost unchanged with the pressure. These characters will decrease the vacuum packaging request of the FPA and should be valuable for practical applications. [ABSTRACT FROM AUTHOR]

Background: Retinitis pigmentosa (RP) is a group of inherited retinal dystrophies characterized by night blindness, progressive peripheral visual field loss, and loss of central vision. Fifty-three RP pathogenic genes are responsible for RP. Pre-mRNA processing factor 31( PRPF31) gene is the third most common cause of autosomal dominant retinitis pigmentosa (adRP), and so far more than 40 mutations in PRPF31 have been detected. Purpose: To identify the underlying genetic defect in a five-generation Chinese family affected with adRP and to study the genotype-phenotype relationship of this family. Methods: Detailed clinical investigations were undertaken and peripheral blood samples were collected from 25 individuals. Microsatellite (STR) markers tightly linked to genes known to be responsible for adRP were selected for linkage analysis. Exons and adjacent splice junctions of the candidate gene were amplified and sequenced. Results: This adRP family exhibited an incomplete penetrance of the RP phenotype. In affected individuals, age of disease onset was from infancy to 4 years of age. Typical RP features were associated with this mutation. Linkage analysis identified a maximum two-point LOD score of 3.20 with D19S418, which is close to PRPF31. A mutation PRPF31: (c.358-359 del AA) was identified by linkage analysis. Conclusions: A PRPF31 mutation was identified to be responsible for adRP in a large Chinese family. Our findings expand the mutation spectrum of RP in the Chinese population. [ABSTRACT FROM AUTHOR]