Fine-grained classification of cervical cells into different abnormality levels is of great clinical importance but remains very challenging. Contrary to traditional classification methods that rely on handcrafted or engineered features, convolution neural network (CNN) can classify cervical cells based on automatically learned deep features. However, CNN in previous studies do not involve cell morphological information, and it is unknown whether morphological features can be directly modeled by CNN to classify cervical cells. This paper presents a CNN-based method that combines cell image appearance with cell morphology for classification of cervical cells in Pap smear. The training cervical cell dataset consists of adaptively re-sampled image patches coarsely centered on the nuclei. Several CNN models (AlexNet, GoogleNet, ResNet and DenseNet) pre-trained on ImageNet dataset were fine-tuned on the cervical dataset for comparison. The proposed method is evaluated on the Herlev cervical dataset by five-fold cross-validation at patient level splitting. Results show that by adding cytoplasm and nucleus masks as raw morphological information into appearance-based CNN learning, higher classification accuracies can be achieved in general. Among the four CNN models, GoogleNet fed with both morphological and appearance information obtains the highest classification accuracies of 94.5% for 2-class classification task and 64.5% for 7-class classification task. Our method demonstrates that combining cervical cell morphology with appearance information can provide improved classification performance, which is clinically important for early diagnosis of cervical dysplastic changes.
It is required to calculate the stored reactive energy of an antenna in order to evaluate its Q factor. Although it has been investigated for a long time, some issues still need further clarification. The main difficulty involved is that the reactive energy of an antenna tends to become infinitely large when integrating the conventionally defined energy density in the whole space outside a small sphere containing the antenna. The reactive energy is usually made to be bounded by subtracting an additional term associated with the radiation fields. However, there exists no well-accepted accurate definition for this additional term that is valid for all cases. By rechecking the definition of reactive energies, an alternative formulation is proposed in this paper which can separate the reactive energy and the radiation energy explicitly based on source-potentials. The clearly defined reactive energy is bounded without necessity to subtract that additional term, and the resultant formulae are easy to implement.
The ORCID identification number(s) for the author(s) of this article can be found under https://doi.org/10.1002/adem.202101080.
Human recombinant Fab fragments specific for the spike protein of severe acute respiratory syndrome coronavirus (SARS-CoV) were screened from a human Fab library, which was generated from RNAs from peripheral lymphocytes of convalescent SARS patients. Among 50 randomly picked clones, 12 Fabs specially reacted with S protein by an enzyme-linked immunosorbent assay. The microneutralizing test showed that one clone, designated M1A, had neutralizing activity on Vero E6 cells against SARS-CoV. DNA sequence analysis indicated that the light-and heavy-chain genes of M1A Fab belong to the 2a and 4f families, respectively. A neutralizing test on purified M1A demonstrated that 0.5 mg/ml of M1A completely inhibited SARS-CoV activity, with an absence of cytopathic effect for 7 days. Real-time fluorescence reverse transcription-PCR also proved the neutralizing capacity of M1A. These data showed that the number of virus copies was significantly reduced in the M1A-treated group, suggesting an important role for M1A in passive immunoprophylaxis against the SARS virus.During 2002 to 2003, severe acute respiratory syndrome (SARS) broke out worldwide. There is no effective medicine to cure this disease. Previous studies showed SARS patients could make great progress if they were given serum from convalescent SARS patients, indicating that generation of a human monoclonal antibody (MAb) against SARS coronavirus (SARS-CoV) may be helpful for passive immunoprophylaxis against SARS virus (8). The spike protein (S protein) of SARS-CoV is an important target for neutralizing antibody (9). We constructed a human antibody library by the phage display technique and used the S protein of SARS-CoV as the target to screen the phage antibody library. This study describes the generation of one neutralizing SARS-CoV-specific human antibody Fab molecule by panning selection of combinatorial Fab libraries against S proteins and quantitative analysis of the binding and neutralizing activity of the Fab molecules. MATERIALS AND METHODSConstruction of a Fab phage display library. Total RNA was extracted from peripheral blood lymphocytes of convalescent SARS patients (RNA isolation kit; Stratagene), and mRNA was reverse transcribed into cDNA using an oligo(dT) primer (Gibco/BRL). Fd segment (variable and first constant domains) genes and light-chain genes were amplified by using primers specific for the human chain genes (1). Then the amplified chains were cloned into the pComb3 phage display vector as described by Bjorling et al. (4). The ligated products were transformed into Escherichia coli XL1-Blue and resulted in a library of 2 ϫ 10 6 clones. The transformants were expanded into a volume of 2 liters, and the resulting phage was recovered as described previously (4).Enrichment of antigen-binding clones by panning. Microtiter wells were coated overnight at 4°C with 50 l of purified SARS-CoV (Beijing 01 strain) lysate antigen (30 g/ml), and after a blocking step with 3% bovine serum albumin (BSA)-phosphate-buffered saline (PBS), 50 l Fab phages (10 ...
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.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.