PCCT: Progressive Class-Center Triplet Loss for Imbalanced Medical Image Classification.

Imbalanced training data in medical image diagnosis is a significant challenge for diagnosing rare diseases. For this purpose, we propose a novel two-stage Progressive Class-Center Triplet (PCCT) framework to overcome the class imbalance issue. In the first stage, PCCT designs a class-balanced triplet loss to coarsely separate distributions of different classes. Triplets are sampled equally for each class at each training iteration, which alleviates the imbalanced data issue and lays solid foundation for the successive stage. In the second stage, PCCT further designs a class-center involved triplet strategy to enable a more compact distribution for each class. The positive and negative samples in each triplet are replaced by their corresponding class centers, which prompts compact class representations and benefits training stability. The idea of class-center involved loss can be extended to the pair-wise ranking loss and the quadruplet loss, which demonstrates the generalization of the proposed framework. Extensive experiments support that the PCCT framework works effectively for medical image classification with imbalanced training images. On four challenging class-imbalanced datasets (two skin datasets Skin7 and Skin 198, one chest X-ray dataset ChestXray-COVID, and one eye dataset Kaggle EyePACs), the proposed approach respectively obtains the mean F1 score 86.20, 65.20, 91.32, and 87.18 over all classes and 81.40, 63.87, 82.62, and 79.09 for rare classes, achieving state-of-the-art performance and outperforming the widely used methods for the class imbalance issue.

Molecular insight into pentraxin-3: Update advances in innate immunity, inflammation, tissue remodeling, diseases, and drug role.

Pentraxin-3 (PTX3) is the prototype of the long pentraxin subfamily, an acute-phase protein consisting of a C-terminal pentraxin domain and a unique N-terminal domain. PTX3 was initially isolated from human umbilical vein endothelial cells and human FS-4 fibroblasts. It was subsequently found to be also produced by synoviocytes, chondrocytes, osteoblasts, smooth muscle cells, myeloid dendritic cells, epithelial cells, and tumor cells. Various modulatory factors, such as miRNAs, cytokines, drugs, and hypoxic conditions, could regulate the expression level of PTX3. PTX3 is essential in regulating innate immunity, inflammation, angiogenesis, and tissue remodeling. Besides, PTX3 may play dual (pro-tumor and anti-tumor) roles in oncogenesis. PTX3 is involved in the occurrence and development of many non-cancerous diseases, including COVID-19, and might be a potential biomarker indicating the prognosis, activity,and severity of diseases. In this review, we summarize and discuss the potential roles of PTX3 in the oncogenesis and pathogenesis of non-cancerous diseases and potential targeted therapies based on PTX3.

Deep Learning Enables Accurate Diagnosis of Novel Coronavirus (COVID-19) With CT Images.

A novel coronavirus (COVID-19) recently emerged as an acute respiratory syndrome, and has caused a pneumonia outbreak world-widely. As the COVID-19 continues to spread rapidly across the world, computed tomography (CT) has become essentially important for fast diagnoses. Thus, it is urgent to develop an accurate computer-aided method to assist clinicians to identify COVID-19-infected patients by CT images. Here, we have collected chest CT scans of 88 patients diagnosed with COVID-19 from hospitals of two provinces in China, 100 patients infected with bacteria pneumonia, and 86 healthy persons for comparison and modeling. Based on the data, a deep learning-based CT diagnosis system was developed to identify patients with COVID-19. The experimental results showed that our model could accurately discriminate the COVID-19 patients from the bacteria pneumonia patients with an AUC of 0.95, recall (sensitivity) of 0.96, and precision of 0.79. When integrating three types of CT images, our model achieved a recall of 0.93 with precision of 0.86 for discriminating COVID-19 patients from others. Moreover, our model could extract main lesion features, especially the ground-glass opacity (GGO), which are visually helpful for assisted diagnoses by doctors. An online server is available for online diagnoses with CT images by our server (http://biomed.nscc-gz.cn/model.php). Source codes and datasets are available at our GitHub (https://github.com/SY575/COVID19-CT).

Shift in the Dominant Sequence Type of Carbapenem-Resistant Klebsiella pneumoniae Bloodstream Infection from ST11 to ST15 at a Medical Center in Northeast China, 2015-2020.

OBJECTIVE: To investigate the clinical characteristics and molecular epidemiology of carbapenem-resistant Klebsiella pneumoniae (CRKP) bloodstream infection at a medical center in northeast China, especially after coronavirus disease (COVID-19) pandemic. METHODS: Fifty-one patients were diagnosed with CRKP bloodstream infection between January 2015 and December 2020, among which 42 isolates were available for further study. Species identification and antibiotic susceptibilities were tested with matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and VITEK 2 systems. Carbapenemase genes, virulence genes and MLST genes were detected by polymerase chain reaction. Moreover, the string test and serum killing assay were performed to evaluate the virulence of the CRKP isolates. RESULTS: During the six-year period, the detection rate of CRKP in bloodstream infection showed an increasing trend, with the intensive care unit, hematology and respiratory medicine wards mainly affected. Molecular epidemiology analyses showed that KPC-2 was the dominant carbapenemase gene. In addition, the dominant sequence type (ST) of CRKP shifted from ST11 to ST15 strains, which were all sensitive to amikacin in contrast to the ST11 stains. Furthermore, ST15 CRKP strains were positive for the KfuB virulence gene and more resistant to serum killing compared to the ST11 CRKP strains. Nonetheless, the mortality rate of patients infected with ST11 and ST15 CRKP did not show any significant differences. CONCLUSION: A shift in the dominant sequence type of CRKP bloodstream infections from ST11 to ST15 was observed during the years 2015-2020. Compared to ST11, the ST15 CRKP strains showed amikacin sensitivity, positivity for KfuB gene, and serum resistance, which may indicate stronger virulence.

Human endeavor for anti-SARS-CoV-2 pharmacotherapy: A major strategy to fight the pandemic.

The global spread of COVID-19 constitutes the most dangerous pandemic to emerge during the last one hundred years. About seventy-nine million infections and more than 1.7 million death have been reported to date, along with destruction of the global economy. With the uncertainty evolved by alarming level of genome mutations, coupled with likelihood of generating only a short lived immune response by the vaccine injections, the identification of antiviral drugs for direct therapy is the need of the hour. Strategies to inhibit virus infection and replication focus on targets such as the spike protein and non-structural proteins including the highly conserved RNA-dependent-RNA-polymerase, nucleotidyl-transferases, main protease and papain-like proteases. There is also an indirect option to target the host cell recognition systems such as angiotensin-converting enzyme 2 (ACE2), transmembrane protease, serine 2, host cell expressed CD147, and the host furin. A drug search strategy consensus in tandem with analysis of currently available information is extremely important for the rapid identification of anti-viral. An unprecedented display of cooperation among the scientific community regarding SARS-CoV-2 research has resulted in the accumulation of an enormous amount of literature that requires curation. Drug repurposing and drug combinations have drawn tremendous attention for rapid therapeutic application, while high throughput screening and virtual searches support de novo drug identification. Here, we examine how certain approved drugs targeting different viruses can play a role in combating this new virus and analyze how they demonstrate efficacy under clinical assessment. Suggestions on repurposing and de novo strategies are proposed to facilitate the fight against the COVID-19 pandemic.

Novel Coronavirus Polymerase and Nucleotidyl-Transferase Structures: Potential to Target New Outbreaks.

The pandemic outbreak of a new coronavirus (CoV), SARS-CoV-2, has captured the world's attention, demonstrating that CoVs represent a continuous global threat. As this is a highly contagious virus, it is imperative to understand RNA-dependent-RNA-polymerase (RdRp), the key component in virus replication. Although the SARS-CoV-2 genome shares 80% sequence identity with severe acute respiratory syndrome SARS-CoV, their RdRps and nucleotidyl-transferases (NiRAN) share 98.1% and 93.2% identity, respectively. Sequence alignment of six coronaviruses demonstrated higher identity among their RdRps (60.9%-98.1%) and lower identity among their Spike proteins (27%-77%). Thus, a 3D structural model of RdRp, NiRAN, non-structural protein 7 (nsp7), and nsp8 of SARS-CoV-2 was generated by modeling starting from the SARS counterpart structures. Furthermore, we demonstrate the binding poses of three viral RdRp inhibitors (Galidesivir, Favipiravir, and Penciclovir), which were recently reported to have clinical significance for SARS-CoV-2. The network of interactions established by these drug molecules affirms their efficacy to inhibit viral RNA replication and provides an insight into their structure-based rational optimization for SARS-CoV-2 inhibition.