Content-based image retrieval (CBIR) techniques have currently gained increasing popularity in the medical field because they can use numerous and valuable archived images to support clinical decisions. In this paper, we concentrate on developing a CBIR system for retrieving brain tumors in T1-weighted contrast-enhanced MRI images. Specifically, when the user roughly outlines the tumor region of a query image, brain tumor images in the database of the same pathological type are expected to be returned. We propose a novel feature extraction framework to improve the retrieval performance. The proposed framework consists of three steps. First, we augment the tumor region and use the augmented tumor region as the region of interest to incorporate informative contextual information. Second, the augmented tumor region is split into subregions by an adaptive spatial division method based on intensity orders; within each subregion, we extract raw image patches as local features. Third, we apply the Fisher kernel framework to aggregate the local features of each subregion into a respective single vector representation and concatenate these per-subregion vector representations to obtain an image-level signature. After feature extraction, a closed-form metric learning algorithm is applied to measure the similarity between the query image and database images. Extensive experiments are conducted on a large dataset of 3604 images with three types of brain tumors, namely, meningiomas, gliomas, and pituitary tumors. The mean average precision can reach 94.68%. Experimental results demonstrate the power of the proposed algorithm against some related state-of-the-art methods on the same dataset.
Figure 1: Comparison of feature maps of CNN (ResNet-101) [18], Visual Transformer (DeiT-S) [41], and the proposed Conformer. The patch embeddings in transformer are reshaped to feature maps for visualization. While CNN activates discriminative local regions (e.g., the peacock's head in (a) and tail in (e)), the CNN branch of Conformer takes advantage of global cues from the visual transformer and thereby activates complete object (e.g., full extent of the peacock in (b) and (f)). Compared with CNN, local feature details of the visual transformer are deteriorated (e.g., (c) and (g)). In contrast, the transformer branch of Conformer retains the local feature details from CNN while depressing the background (e.g., the peacock contours in (d) and (h) are more complete than those in (c) and (g)).
Pulmonary fibrosis is the end stage of a broad range of heterogeneous interstitial lung diseases and more than 200 factors contribute to it. In recent years, the relationship between virus infection and pulmonary fibrosis is getting more and more attention, especially after the outbreak of SARS-CoV-2 in 2019, however, the mechanisms underlying the virus-induced pulmonary fibrosis are not fully understood. Here, we review the relationship between pulmonary fibrosis and several viruses such as Human T-cell leukemia virus (HTLV), Human immunodeficiency virus (HIV), Cytomegalovirus (CMV), Epstein–Barr virus (EBV), Murine γ-herpesvirus 68 (MHV-68), Influenza virus, Avian influenza virus, Middle East Respiratory Syndrome (MERS)-CoV, Severe acute respiratory syndrome (SARS)-CoV and SARS-CoV-2 as well as the mechanisms underlying the virus infection induced pulmonary fibrosis. This may shed new light on the potential targets for anti-fibrotic therapy to treat pulmonary fibrosis induced by viruses including SARS-CoV-2.
This study aimed to explore the effect and mechanism of H. cordata vapor extract on acute lung injury (ALI) and rapid pulmonary fibrosis (RPF). We applied the volatile extract of HC to an RPF rat model and analyzed the effect on ALI and RPF using hematoxylin-eosin (H&E) staining, routine blood tests, a cell count of bronchoalveolar lavage fluid (BALF), lactate dehydrogenase (LDH) content, van Gieson (VG) staining, hydroxyproline (Hyp) content and the dry/wet weight ratio. The expression of IFN-γ/STAT(1), IL-4/STAT(6) and TGF-β(1)/Smads was analyzed using ELISA, immunohistochemistry and western blotting methods. The active ingredients of the HC vapor extract were analyzed using a gas chromatograph-mass spectrometer (GC-MS), and the effects of the active ingredients of HC on the viability of NIH/3T3 and RAW264.7 cells were detected using an MTT assay. The active ingredients of the HC vapor extract included 4-terpineol, α-terpineol, l-bornyl acetate and methyl-n-nonyl ketone. The results of the lung H&E staining, Hyp content, dry/wet weight ratio and VG staining suggested that the HC vapor extract repaired lung injury and reduced RPF in a dose-dependent manner and up-regulated IFN-γ and inhibited the TGF-β1/Smad pathway in vivo. In vitro, it could inhibit the viability of RAW264.7 and NIH/3T3 cells. It also dose-dependently inhibited the expression of TGF-β1 and enhanced the expression of IFN-γ in NIH/3T3. The HC vapor extract inhibited LPS-induced RPF by up-regulating IFN-γ and inhibiting the TGF-β1/Smad pathway.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.