Unsupervised Domain Adaptation with Asymmetrical Margin Disparity loss and Outlier Sample Extraction.

Unsupervised domain adaptation (UDA) trains models using labeled data from a specific source domain and then transferring the knowledge to certain target domains that have few or no labels. Many prior measurement-based works achieve lots of progress, but their feature distinguishing abilities to classify target samples with similar features are not enough; they do not adequately consider the confusing samples in the target domain that are similar to the source domain; and they don't consider negative transfer of the outlier sample in source domain. We address these issues in our work and propose an UDA method with asymmetrical margin disparity loss and outlier sample extraction, called AMD-Net with OSE. We propose an Asymmetrical Margin Disparity Discrepancy (AMD) method and a training strategy based on sample selection mechanism to make the network have better feature extraction ability and the network gets rid of local optimal. Firstly, in the AMD method, we design a multi-label entropy metric to evaluate the marginal disparity loss of the confusing samples in the target domain. This asymmetric marginal disparity loss designment uses the different entropy measurement algorithms of the two domains to excavate the differences of the two domains as much as possible, so as to find the common features of the two domains. Secondly, A sample selection mechanism is designed to evaluate which part of the sample in target domain is confusable. We define the certainty of the sample in the target domain, adopt a progressive learning scheme, and adopt one-hot marginal disparity loss for most of the samples in the target domain with low uncertainty and easy to distinguish. The multi-label marginal calculation method is used only for the uncertainty samples in the target domain whose certainty is less than the threshold value, so that the network can get rid of the local optimal as much as possible. At last, we further propose an outlier sample extraction algorithm (OSE) based on weighted cosine similarity distance for source domain to reduce the negative migration effect caused by outlier samples in the source domain. Extensive experiments on four datasets Office-31, Office-Home, VisDA-2017 and DomainNet demonstrate that our method works well in various UDA settings and outperforms the state-of-the-art methods.

A Lightweight and Robust Framework for Circulating Genetically Abnormal Cells (CACs) Identification Using 4-Color Fluorescence In Situ Hybridization (FISH) Image and Deep Refined Learning.

Circulating genetically abnormal cells (CACs) constitute an important biomarker for cancer diagnosis and prognosis. This biomarker offers high safety, low cost, and high repeatability, which can serve as a key reference in clinical diagnosis. These cells are identified by counting fluorescence signals using 4-color fluorescence in situ hybridization (FISH) technology, which has a high level of stability, sensitivity, and specificity. However, there are some challenges in CACs identification, due to the difference in the morphology and intensity of staining signals. In this concern, we developed a deep learning network (FISH-Net) based on 4-color FISH image for CACs identification. Firstly, a lightweight object detection network based on the statistical information of signal size was designed to improve the clinical detection rate. Secondly, the rotated Gaussian heatmap with a covariance matrix was defined to standardize the staining signals with different morphologies. Then, the heatmap refinement model was proposed to solve the fluorescent noise interference of 4-color FISH image. Finally, an online repetitive training strategy was used to improve the model's feature extraction ability for hard samples (i.e., fracture signal, weak signal, and adjacent signals). The results showed that the precision was superior to 96%, and the sensitivity was higher than 98%, for fluorescent signal detection. Additionally, validation was performed using the clinical samples of 853 patients from 10 centers. The sensitivity was 97.18% (CI 96.72-97.64%) for CACs identification. The number of parameters of FISH-Net was 2.24 M, compared to 36.9 M for the popularly used lightweight network (YOLO-V7s). The detection speed was about 800 times greater than that of a pathologist. In summary, the proposed network was lightweight and robust for CACs identification. It could greatly increase the review accuracy, enhance the efficiency of reviewers, and reduce the review turnaround time during CACs identification.

Attention Mask R-CNN with edge refinement algorithm for identifying circulating genetically abnormal cells.

Recent studies have suggested that circulating tumor cells with abnormalities in gene copy numbers in mononuclear cell-enriched peripheral blood samples, such as circulating genetically abnormal cells (CACs), can be used as a non-invasive tool to detect patients with benign pulmonary nodules. These cells are identified through fluorescence signals counting by using 4-color fluorescence in situ hybridization (FISH) technology that exhibits high stability, sensitivity, and specificity. When FISH data are analyzed, the overlapping cells and fluorescence noise is a great challenge for identifying of CACs, thereby seriously affecting the efficiency of clinical diagnosis. To address this problem, in this study, we proposed an end-to-end FISH-based method (CACNET) for CAC identification. CACNET achieved nuclear segmentation and counted 4-color staining signals through improved Mask region-based convolutional neural network (R-CNN), followed by cell category (normal cell, deletion cell, gain cell, or CAC) according to pathological criteria. Firstly, the segmentation accuracy of overlapping nuclei was improved by adding an edge constraint head during training. Then, the interference of fluorescence noise was reduced by fusing non-local module to reconstruct the feature extraction network of Mask R-CNN. We trained and tested the proposed model on a dataset comprising 700 frames with 58,083 nuclei. The Accuracy, Sensitivity, and Specificity (overall performance metric for the algorithm) of CAC identification with CACNET were 94.06%, 92.1%, and 99.8%, respectively. Moreover, the developed method exhibited approximately identification speed of approximately 0.22 s per frames. The results showed that the proposed method outperformed the existing CAC identification methods, making it a promising approach for early screening of lung cancer.

Multi-attention representation network partial domain adaptation for COVID-19 diagnosis.

The outbreak of COVID-19 threatens the safety of all human beings. Rapid and accurate diagnosis of patients is the effective way to prevent the rapid spread of COVID-19. The current computer-aided diagnosis of COVID-19 requires extensive labeled data for training, and this undoubtedly increases human and material resources costs. Domain adaptation (DA), an existing promising approach, can transfer knowledge from rich labeled pneumonia datasets for COVID-19 diagnosis and classification. However, due to the differences in feature distribution and task semantic between pneumonia and COVID-19, negative transfer may reduce the performance in diagnosis COVID-19 and pneumonia. Furthermore, the training data is usually mixed with many noise samples in practice, and this also poses new challenges for domain adaptation. As a kind of domain adaptation, partial domain adaptation (PDA) can well avoid outlier samples in the source domain and achieve good classification performance in the target domain. However, the existing PDA methods all learn a single feature representation; this can only learn local information about the inputs and ignore other important information in the samples. Therefore multi-attention representation network partial domain adaptation (MARPDA) is proposed in this paper to overcome the above shortcomings of PDA. In MARPDA, we construct the multiple representation networks with attention to acquire the image representation and effectively learn knowledge from different feature spaces. We design the sample-weighted strategy to achieve partial data transfer and address the negative transfer of noise data during training. MARPDA adapts to complex application scenarios and learns fine-grained features of the image from multiple representations. We apply the model to classify pneumonia and COVID-19 respectively, and evaluate it in qualitative and quantitative manners. The experimental results show that our classification accuracy is higher than that of the existing state-of-the-art methods. The stability and reliability of the proposed method are validated by the confusion matrix and the performance curves experiments. In summary, our method has better performance for diagnosis COVID-19 compared to the existing state-of-the-art methods.

An efficient fluorescence in situ hybridization (FISH)-based circulating genetically abnormal cells (CACs) identification method based on Multi-scale MobileNet-YOLO-V4.

Background: Circulating tumor cells (CTCs) acting as "liquid biopsy" of cancer are cells that have been shed from the primary tumor, which cause the development of a secondary tumor in a distant organ site, leading to cancer metastasis. Recent research suggests that CTCs with abnormalities in gene copy numbers in mononuclear cell-enriched peripheral blood samples, namely circulating genetically abnormal cells (CACs), could be used as a non-invasive decision tool to detect patients with benign pulmonary nodules. Such cells are identified by counting the fluorescence signals of fluorescence in situ hybridization (FISH). However, owing to the rarity of CACs in the blood, identification of CACs using this technique is time-consuming and is a drawback of this method. Methods: This study has proposed an efficient and automatic FISH-based CACs identification approach which is based on a combination of the high accuracy of You Only Look Once (YOLO)-V4 and the lightweight and rapidness of MobileNet-V3. The backbone of YOLO-V4 was replaced with MobileNet-V3 to improve the detection efficiency and prevent overfitting, and the architecture of YOLO-V4 was optimized by utilizing a new feature map with a larger scale to enable the enhanced detection ability for small targets. Results: We trained and tested the proposed model using a dataset containing more than 7,000 cells based on five-fold cross-validation. All the images in the dataset were 2,448×2,048 (pixels) in size. The number of cells in each image was >70. The accuracy of four-color fluorescence signals detection for our proposed model were all approximately 98%, and the mean average precision (mAP) were close to 100%. The final outcome of the developed method was the type of cells, i.e., normal cells, CACs, gaining cells or deletion cells. The method had a CACs identification accuracy of 93.86% (similar to an expert pathologist), and a detection speed that was about 500 times greater than that of a pathologist. Conclusions: The developed method could greatly increase the review accuracy, enhance the efficiency of reviewers, and reduce the review turnaround time during CACs identification.

Circulating Genetically Abnormal Cells Add Non-Invasive Diagnosis Value to Discriminate Lung Cancer in Patients With Pulmonary Nodules ≤10 mm.

BACKGROUND: Lung cancer screening using low-dose computed tomography (LDCT) often leads to unnecessary biopsy because of the low specificity among patients with pulmonary nodules ≤10 mm. Circulating genetically abnormal cells (CACs) can be used to discriminate lung cancer from benign lung disease. To examine the diagnostic value of CACs in detecting lung cancer for patients with malignant pulmonary nodules ≤10 mm. METHODS: In this prospective study, patients with pulmonary nodules ≤10 mm who were detected at four hospitals in China from January 2019 to January 2020 were included. CACs were detected using fluorescence in-situ hybridization. All patients were confirmed as lung cancer or benign disease by further histopathological examination. Multivariable logistic regression models were established to detect the presence of lung cancer using CACs and other associated characteristics. Receiver operating characteristic analysis was used to evaluate the performance of CACs for lung cancer diagnosis. RESULTS: Overall, 125 patients were included and analyzed. When the cutoff value of CACs was >2, the sensitivity and specificity for lung cancer were 70.5 and 86.4%. Male (OR = 0.330, P = 0.005), maximum solid nodule (OR = 2.362, P = 0.089), maximum nodule located in upper lobe (OR = 3.867, P = 0.001), and CACs >2 (OR = 18.525, P < 0.001) met the P < 0.10 criterion for inclusion in the multivariable models. The multivariable logistic regression model that included the dichotomized CACs (>2 vs. ≤2) and other clinical factors (AUC = 0.907, 95% CI = 0.842-0.951) was superior to the models that only considered dichotomized CACs or other clinical factors and similar to the model with numerical CACs and other clinical factors (AUC = 0.913, 95% CI = 0.850-0.956). CONCLUSION: CACs presented a significant diagnostic value in detecting lung cancer for patients with pulmonary nodules ≤10 mm.

Danhong injection in the treatment of idiopathic pulmonary fibrosis: A protocol for systematic review and meta-analysis.

BACKGROUND: Many studies have reported that the effects of danhong injection on idiopathic pulmonary fibrosis. However, its effects are still not well understood. The aim of this study is to assess the effects of danhong injection in the management of idiopathic pulmonary fibrosis. METHODS: Electronic databases such as PubMed, MEDLINE, EMbase, Web of Science, Cochrane Library, China National Knowledge Infrastructure, WanFang, the Chongqing VIP Chinese Science and Technology Periodical Database, and China biomedical literature database will be searched without limitations of language and geographical location. Two researchers will independently conduct research selection, data extraction, and research quality assessment. The RevMan 5.3 software and Stata 14.0 software are used for statistical analysis. RESULTS: This study will provide high-quality comprehensive evidence for the effectiveness and safety of danhong injection in the treatment of idiopathic pulmonary fibrosis. CONCLUSIONS: The results obtained from this study will define the basis for the effectiveness and safety of danhong injection in the treatment of idiopathic pulmonary fibrosis.

Detection of circulating genetically abnormal cells in peripheral blood for early diagnosis of non-small cell lung cancer.

BACKGROUND: Circulating genetically abnormal cells (CACs) with specific chromosome variations have been confirmed to be present in non-small cell lung cancer (NSCLC). However, the diagnostic performance of CAC detection remains unclear. This study aimed to evaluate the potential clinical application of the CAC test for the early diagnosis of NSCLC. METHODS: In this prospective study, a total of 339 participants (261 lung cancer patients and 78 healthy volunteers) were enrolled. An antigen-independent fluorescence in situ hybridization was used to enumerate the number of CACs in peripheral blood. RESULTS: Patients with early-stage NSCLC were found to have a significantly higher number of CACs than those of healthy participants (1.34 vs. 0.19; P < 0.001). The CAC test displayed an area under the receiver operating characteristic (ROC) curve of 0.76139 for discriminating stage I NSCLC from healthy participants with 67.2% sensitivity and 80.8% specificity, respectively. Compared with serum tumor markers, the sensitivity of CAC assays for distinguishing early-stage NSCLC was higher (67.2% vs. 48.7%, P < 0.001), especially in NSCLC patients with small nodules (65.4% vs. 36.5%, P = 0.003) and ground-glass nodules (pure GGNs: 66.7% vs. 40.9%, P = 0.003; mixed GGNs: 73.0% vs. 43.2%, P < 0.001). CONCLUSIONS: CAC detection in early stage NSCLC was feasible. Our study showed that CACs could be used as a promising noninvasive biomarker for the early diagnosis of NSCLC. KEY POINTS: What this study adds: This study aimed to evaluate the potential clinical application of the CAC test for the early diagnosis of NSCLC. Significant findings of the study: CAC detection in early stage NSCLC was feasible. Our study showed that CACs could be used as a promising noninvasive biomarker for the early diagnosis of NSCLC.

Sensitive detection of atrazine in tap water using TELISA.

A highly sensitive flow injection analysis (FIA)-based thermal enzyme-linked immunoassay, TELISA, was developed for the rapid detection of atrazine (ATZ) in tap water. ATZ and ß-lactamase-labeled ATZ were employed in a competitive immunoassay using a monoclonal antibody (mAb). After the off-column liquid-phase competition, the mAb was captured on the Protein G Sepharose™ 4 Fast Flow (PGSFF) column support material. Injected ß-lactamase substrate ampicillin was degraded by the column-bound ATZ-ß-lactamase, generating a detectable heat signal. Several assay parameters were optimized, including substrate concentration, flow rates and regeneration conditions, as well as the mAb and ATZ-ß dilution ratios and concentrations. The assay linear range was 0.73-4.83 ng mL(-1) with a detection limit of 0.66 ng mL(-1). An entire heat signal requires 10 min for generation, and the cycle time is less than 40 min. The results were reproducible and stable. ATZ-spiked tap water samples exhibited a recovery rate of 103%-116%, which correlated with the UHPLC-MS/MS measurements. We attributed this significant increase in sensitivity over our previously published work to the following factors: (i) the capture of already-formed immune complexes on the column via immobilized Protein G, which eliminated chemical immobilization of the antibody; (ii) off-column preincubation allows the formation of immune complexes under nearly ideal conditions; and (iii) multiple buffers can be used to, in one case, enhance immune-complex formation and in the other to maximize enzymatic activity. Furthermore, the scheme creates a universal assay platform in which sensing is performed in the off-column incubation and detection after capture in the enzyme thermistor (ET) detector, which opens up the possibility of detecting any antigen for which antibodies were available.

Selection of bisphenol A - single-chain antibodies from a non-immunized mouse library by ribosome display.

Developing reagents with high affinity and specificity are critical to detect the environmental hormones or toxicants. Ribosome display technology has been widely used in functional protein or peptide screening and in directed evolution of protein molecules in vitro. In this study, single-chain variable fragments (scFvs) against bisphenol A (BPA) were selected from a library constructed from splenocytes of non-immunized mice. After five rounds of selection, the selected scFvs bound to BPA with high affinity. Indirect competitive enzyme-linked immunosorbent assay (ELISA) was introduced to screen the antibody affinity and specificity to BPA. The equilibrium dissociation constants (KDS) of one clone was 1.76µM as determined by surface plasmon resonance (SPR). This study indicated that ribosome display can isolate binders to small molecules from a non-immunized naive library without any in vivo steps and can generate recombinant antibodies efficiently and rapidly. In addition, this study provides a methodological framework for detection of small molecules using recombinant antibodies.

Application of suspension array for simultaneous detection of four different mycotoxins in corn and peanut.

Mycotoxins are highly toxic contaminants in foodstuffs and feedstuffs. The study presents a novel suspension array technology for quantifying four mycotoxins, namely, aflatoxin B1, deoxynivalenol, T-2 toxin, and zearalenone, in corn and peanut. Using 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride, the complete antigens of the mycotoxins became attached to the microspheres with viable activity. The optimal concentrations of each antibody and biotin-rabbit anti-goat IgG were obtained through chessboard titration. The four mycotoxins were detected simultaneously and quantitatively in corn and peanut using indirect competitive immunoassay. Multi-channel standard curves with appropriate logistic correlation (R(2)>0.9819) were respectively plotted. The broad working ranges with three to four orders of magnitude were calculated, and limits of detection at the pg level were found to be better than those obtained using high-performance liquid chromatography. The recovery rates in the actual samples generally ranged from 80.16% to 117.65%, with an intra-assay coefficient of variation lower than 15%, which indicated high accuracy and repeatability. A suspension array method for the simultaneous detection of the four mycotoxins within 4h was successfully developed using minimal samples; the method was proven to have high throughput, flexibility, accuracy and reproducibility. The approach could detect multiple contaminants in actual samples.

Single-wall carbon nanotubes induce oxidative stress in rat aortic endothelial cells.

Oxidative stress is a major factor contributing to endothelial cell damage. Single-wall carbon nanotubes (SWCNTs) have oxidative properties; however, the oxidative effects of SWCNTs on endothelial cells are not fully understood. In the present study, we investigated the effects of oxidative stress induced by SWCNTs on rat aortic endothelial cells (RAECs). Various markers of cellular damage were assessed, such as biochemical and ES immunity indexes, and DNA and protein damage. Our findings suggest that RAEC endured oxidative damage following SWCNT exposure. Specifically, after SWCNTs exposure, non-enzymatic antioxidant glutathione was activated prior to superoxide dismutase activation in order to defend against oxidative stress. Additionally, it was found that as SWCNT concentration increased, so did the stress protein, heme oxygenase-1 (HO-1), expression levels. These changes may induce RAEC damage, and result in many serious diseases.

Construction of ribosome display library based on lipocalin scaffold and screening anticalins with specificity for estradiol.

A new anticalin against estradiol (E(2)), a kind of endocrine disruptor, was obtained in the present study to detect E(2) levels. A member of the lipocalin family from Pieris brassicae called bilin-binding protein (BBP) was employed for the preparation of a random library to specifically complex E(2). Sixteen amino acid residues at the center of the binding site, which were formed by four loops on top of an eight-stranded ß-barrel, were subjected to targeted random mutagenesis. Estradiol-binding BBP variants so-called 'anticalins', which exhibit binding activity for compounds, such as E(2), were selected from the resulting library by combining both ribosome display and screening techniques. Four variants of complex E(2) with high affinity were identified. These variants exhibited dissociation constants (KDs) as low as 54.265 nM. ELISA showed that ribosome displayed anticalin (E(2)-A) specifically bound E(2). The 50% inhibition concentration (IC(50)) for E(2) was 50 ng mL(-1) and the limit of detection (LOD:IC(10)) was 0.071 ng mL(-1). The experimental results suggest that E(2)-A can be used as a potential anticalin to detect E(2) in animals.

Selection of diethylstilbestrol-specific single-chain antibodies from a non-immunized mouse ribosome display library.

Single chain variable fragments (scFvs) against diethylstilbestrol (DES) were selected from the splenocytes of non-immunized mice by ribosome display technology. A naive library was constructed and engineered to allow in vitro transcription and translation using an E. coli lysate system. Alternating selection in solution and immobilization in microtiter wells was used to pan mRNA-ribosome-antibody (ARM) complexes. After seven rounds of ribosome display, the expression vector pTIG-TRX containing the selected specific scFv DNAs were transformed into Escherichia coli BL21 (DE3) for expression. Twenty-six positive clones were screened and five clones had high antibody affinity and specificity to DES as evidenced by indirect competitive ELISA. Sequence analysis showed that these five DES-specific scFvs had different amino acid sequences, but the CDRs were highly similar. Surface plasmon resonance (SPR) analysis was used to determine binding kinetics of one clone (30-1). The measured K(D) was 3.79 µM. These results indicate that ribosome display technology can be used to efficiently isolate hapten-specific antibody (Ab) fragments from a naive library; this study provides a methodological framework for the development of novel immunoassays for multiple environmental pollutants with low molecular weight detection using recombinant antibodies.

Correction: Selection of Diethylstilbestrol-Specific Single-Chain Antibodies from a Non-Immunized Mouse Ribosome Display Library.

[This corrects the article DOI: 10.1371/journal.pone.0033186.].

Single-walled carbon nanotube induction of rat aortic endothelial cell apoptosis: Reactive oxygen species are involved in the mitochondrial pathway.

The use of nano-sized materials offers exciting new options in technical and medical applications. Single-walled carbon nanotubes are emerging as technologically important in different industries. However, adverse effects on cells have been reported and this may limit their use. We previously found that 200µg/mL of single-walled carbon nanotubes induce apoptosis in rat aorta endothelial cells. The current study aimed to determine the signaling pathway involved in this process. We found that reactive oxygen species generation was involved in activation of the mitochondria-dependent apoptotic pathway. The finding of apoptosis was supported by a number of morphological and biochemical hallmarks, including chromatin condensation, internucleosomal DNA fragmentation, and caspase-3 activation. In conclusion, our results demonstrate that single-walled carbon nanotubes induce apoptosis in rat aorta endothelial cells and that reactive oxygen species are involved in the mitochondrial pathway.