DRA-Net: Medical image segmentation based on adaptive feature extraction and region-level information fusion.

Medical image segmentation is a key task in computer aided diagnosis. In recent years, convolutional neural network (CNN) has made some achievements in medical image segmentation. However, the convolution operation can only extract features in a fixed size region at a time, which leads to the loss of some key features. The recently popular Transformer has global modeling capabilities, but it does not pay enough attention to local information and cannot accurately segment the edge details of the target area. Given these issues, we proposed dynamic regional attention network (DRA-Net). Different from the above methods, it first measures the similarity of features and concentrates attention on different dynamic regions. In this way, the network can adaptively select different modeling scopes for feature extraction, reducing information loss. Then, regional feature interaction is carried out to better learn local edge details. At the same time, we also design ordered shift multilayer perceptron (MLP) blocks to enhance communication within different regions, further enhancing the network's ability to learn local edge details. After several experiments, the results indicate that our network produces more accurate segmentation performance compared to other CNN and Transformer based networks.

CircHAS2 activates CCNE2 to promote cell proliferation and sensitizes the response of colorectal cancer to anlotinib.

BACKGROUND: Tyrosine kinase inhibitors (TKIs) are crucial in the targeted treatment of advanced colorectal cancer (CRC). Anlotinib, a multi-target TKI, has previously been demonstrated to offer therapeutic benefits in previous studies. Circular RNAs (circRNAs) have been implicated in CRC progression and their unique structural stability serves as promising biomarkers. The detailed molecular mechanisms and specific biomarkers related to circRNAs in the era of targeted therapies, however, remain obscure. METHODS: The whole transcriptome RNA sequencing and function experiments were conducted to identify candidate anlotinib-regulated circRNAs, whose mechanism was confirmed by molecular biology experiments. CircHAS2 was profiled in a library of patient-derived CRC organoids (n = 22) and patient-derived CRC tumors in mice. Furthermore, a prospective phase II clinical study of 14 advanced CRC patients with anlotinib-based therapy was commenced to verify drug sensitivity (ClinicalTrials.gov identifier: NCT05262335). RESULTS: Anlotinib inhibits tumor growth in vitro and in vivo by downregulating circHAS2. CircHAS2 modulates CCNE2 activation by acting as a sponge for miR-1244, and binding to USP10 to facilitate p53 nuclear export as well as degradation. In parallel, circHAS2 serves as a potent biomarker predictive of anlotinib sensitivity, both in patient-derived organoids and xenograft models. Moreover, the efficacy of anlotinib inclusion into the treatment regimen yields meaningful clinical responses in patients with high levels of circHAS2. Our findings offer a promising targeted strategy for approximately 52.9% of advanced CRC patients who have high circHAS2 levels. CONCLUSIONS: CircHAS2 promotes cell proliferation via the miR-1244/CCNE2 and USP10/p53/CCNE2 bidirectional axes. Patient-derived organoids and xenograft models are employed to validate the sensitivity to anlotinib. Furthermore, our preliminary Phase II clinical study, involving advanced CRC patients treated with anlotinib, confirmed circHAS2 as a potential sensitivity marker.

RFE-UNet: Remote Feature Exploration with Local Learning for Medical Image Segmentation.

Although convolutional neural networks (CNNs) have produced great achievements in various fields, many scholars are still exploring better network models, since CNNs have an inherent limitation-that is, the remote modeling ability of convolutional kernels is limited. On the contrary, the transformer has been applied by many scholars to the field of vision, and although it has a strong global modeling capability, its close-range modeling capability is mediocre. While the foreground information to be segmented in medical images is usually clustered in a small interval in the image, the distance between different categories of foreground information is uncertain. Therefore, in order to obtain a perfect medical segmentation prediction graph, the network should not only have a strong learning ability for local details, but also have a certain distance modeling ability. To solve these problems, a remote feature exploration (RFE) module is proposed in this paper. The most important feature of this module is that remote elements can be used to assist in the generation of local features. In addition, in order to better verify the feasibility of the innovation in this paper, a new multi-organ segmentation dataset (MOD) was manually created. While both the MOD and Synapse datasets label eight categories of organs, there are some images in the Synapse dataset that label only a few categories of organs. The proposed method achieved 79.77% and 75.12% DSC on the Synapse and MOD datasets, respectively. Meanwhile, the HD95 (mm) scores were 21.75 on Synapse and 7.43 on the MOD dataset.

[Echinacoside hampers malignant progression of breast cancer MCF-7 cells by modulating AKR1B10/ERK signal transduction].

This study analyzes the impact of echinacoside(ECH) in the proliferation, metastasis and adriamycin(ADR) resistance of breast cancer(BC) MCF-7 cells via the modulation of aldo-keto reductase family 1 member 10(AKR1B10)/extracellular signal-regulated kinase(ERK) pathway. The chemical structure of ECH was firstly confirmed. MCF-7 cells were treated with different concentration(0, 10, 20, 40 µg·mL~(-1)) of ECH for 48 h. Western blot was used to analyze expression of AKR1B10/ERK pathway-associated proteins and cell counting kit-8(CCK-8) assay to determine cell viability. MCF-7 cells were collected and classified into control group, ECH group, ECH + Ov-NC group, and ECH + Ov-AKR1B10 group. Then Western blot was employed to analyze the expression of AKR1B10/ERK pathway-associated proteins. CCK-8 and 5-ethynyl-2'-deoxyuridine(EdU) assay were used to examine cell proliferation. Cell migration was appraised with scratch assay, Transwell assay, and Western blot. Eventually, MCF-7 cells were treated with ADR for 48 h to induce ADR resistance. Cell viability was tested by CCK-8 assay and cell apoptosis was estimated based on terminal-deoxynucleoitidyl transferase mediated nick end labeling(TUNEL) assay and Western blot. Based on Protein Data Bank(PDB) and molecular docking, the binding affinity of ECH to AKR1B10 was assessed. Various doses of ECH decreased the expression of AKR1B10/ERK pathway-associated proteins in a dose-dependent manner and declined cell viability compared with the control group. Compared with the control group, 40 µg·mL~(-1) ECH blocked the AKR1B10/ERK pathway in MCF-7 cells and inhibited the proliferation, metastasis and ADR resistance of the cells. Compared with the ECH + Ov-NC group, ECH + Ov-AKR1B10 group showed the recovery of some biological behaviors of MCF-7 cells. ECH also targeted AKR1B10. ECH can inhibit the proliferation, metastasis, and ADR resistance of BC cells by blocking AKR1B10/ERK pathway.

Effect of FOXP2 transcription factor on immune infiltration of thyroid cancer and its potential clinical value.

Background: The clinical outcomes are not always favorable in certain thyroid cancer patients. The effect of Forkhead-box family on immune cells infiltration and tumor microenvironment in thyroid cancer was explored. The role of FOXP2 in tumor invasion and recurrence was investigated consequently. Methods: TIMER and GEPIA were firstly employed to compare FOXPs expression in normal and cancer tissues from multiple human cancers. The results from database were confirmed by quantitative Real Time-PCR and Western blot in matched thyroid cancer and adjacent normal tissues, in addition to a panel of thyroid cancer cell lines and normal thyroid cell. GEPIA platform was employed to discover the possibility of FOXPs as prognostic indicator. TISIBD and UACLCAN were then employed to estimate the influence of FOXPs on lymph node metastasis and tumor staging. GEPIA analysis was initially employed to analyze correlation of FOXPs and tumor immune infiltrating cells, and TIMER dataset was then included for standardization according to tumor purity. Result: Different member of FOXPs showed divergence in expression in various cancer tissues. Lower FOXP1, FOXP2 and higher FOXP3, FOXP4 levels could be identified in thyroid cancer tissues when compared with matched normal tissue. There was an inverse correlation between FOXP2, FOXP4 and immune invasion, whereas FOXP1 and FOXP3 were positively correlated. FOXPs showed remarkable correlations with multiply immune cells. More importantly, only FOXP2 showed the significant effect on recurrence and tumor staging. Conclusion: As immune regulatory factor, the reduction of FOXP2 may affect tumor microenvironments and immune cells infiltration, enhance tumor immune escape, and promote recurrence of thyroid cancer. FOXP2 could be a new potential diagnostic and prognostic marker.

MHANet: A hybrid attention mechanism for retinal diseases classification.

With the increase of patients with retinopathy, retinopathy recognition has become a research hotspot. In this article, we describe the etiology and symptoms of three kinds of retinal diseases, including drusen(DRUSEN), choroidal neovascularization(CNV) and diabetic macular edema(DME). In addition, we also propose a hybrid attention mechanism to classify and recognize different types of retinopathy images. In particular, the hybrid attention mechanism proposed in this paper includes parallel spatial attention mechanism and channel attention mechanism. It can extract the key features in the channel dimension and spatial dimension of retinopathy images, and reduce the negative impact of background information on classification results. The experimental results show that the hybrid attention mechanism proposed in this paper can better assist the network to focus on extracting thr fetures of the retinopathy area and enhance the adaptability to the differences of different data sets. Finally, the hybrid attention mechanism achieved 96.5% and 99.76% classification accuracy on two public OCT data sets of retinopathy, respectively.

Performance Testing and Rapid Solidification Behavior of Stainless Steel Powders Prepared by Gas Atomization.

Gas atomization is a widely used method to produce the raw powder materials for additive manufacturing (AM) usage. After the metal alloy is melted to fusion, gas atomization involves two relatively independent processes: liquid breakup and droplet solidification. In this paper, the solidification behavior of powder during solidification is analyzed by testing the powder's properties and observing microstructure of a martensitic stainless steel (FeCrNiBSiNb). The powder prepared by gas atomization has high sphericity and smooth surface, and the yield of qualified fine powder is 35%. The powder has typical rapid solidification structure. Collision between powders not only promotes nucleation, but also produces more satellite powder. The segregation of elements in powder is smaller as the result of high cooling rate which can reaches 4.2 × 105 K/s in average. Overall, the powder prepared by gas atomization is found to have good comprehensive properties, desired microstructure, and accurate chemical component, and it is suitable for various additive manufacturing techniques.