ABSTRACT
The clustering of large numbers of heterogeneous features is a hot topic in multi-view communities. Most existing multi-view clustering (MvC) methods employ matrix factorization or anchor strategies to handle large-scale datasets. The former operates on the original data and is, therefore, sensitive to noise and feature redundancy, which is reflected in the final clustering performance. The latter requires post -processing steps to generate the clustering results, which may be suboptimal owing to the isolation steps. To address the above problems, we propose one-stage multi-view subspace clustering with dictionary learning (OSMvSC). Specifically, we integrate dictionary learning, representation coefficient matrix learning, and matrix factorization as a unified learning framework, which directly learns the dictionary and representation coefficient matrix to encode the original multi-view data, and obtains the clustering results with linear time complexity without any postprocessing step. By manipulating the class centroid with the nuclear norm, a more compact and discriminative class centroid representation can be obtained to further improve clustering performance. An effective optimization algorithm with guaranteed convergence is designed to solve the proposed method. Substantial experiments on various real-world multi-view datasets demonstrate the effectiveness and superiority of the proposed method. The source code is available at https://github.com/justcallmewilliam/OSMvSC.(c) 2022 Elsevier B.V. All rights reserved.
ABSTRACT
Radiation therapy (RT) to doses of 24-30 Gy is used for the treatment of indolent B-cell lymphoma (BCL);however, significant acute and late ocular effects are common. We aimed to develop a response adapted (RA) strategy that maintains excellent disease outcomes but reduces orbital morbidity. We performed a phase II prospective study of a RA strategy in 50 patients (pts) with stage I-IV orbital indolent BCL. Pts were treated with ultra-low dose (ULD) RT to 4 Gy in 2 fractions and assessed in 3-month intervals for response. Pts with persistent orbital lymphoma were offered an additional 20 Gy in 10 fractions. Pts that had a complete response (CR) to ULD RT were observed. We also evaluated this treatment strategy in a separate 55 pt retrospective cohort. From July 2015-January 2021 51 pts were enrolled. Fifty evaluable pts had follow-up for study inclusion. The median age was 63 years (29-88);62% were female (n=31). Pts had MALT lymphoma (n=32, 64%), follicular lymphoma (FL, n=16, 32%) and low grade BCL (n=6, 12%). Most pts (62%, n=31) had stage I disease limited to one (n=28) or both (n=3) orbits. Pts had newly diagnosed (n=36, 72%);relapsed (n=9, 18%) and refractory lymphoma (n=5, 10%). At a median follow up of 35 months [95% CI 22.2 – 37.4], 90% of pts (n=45) experienced a CR to RA RT, including 44 pts that had a CR to ULD RT (median time to CR 3.4 months) and 1 pt that had a CR after an additional 20 Gy. No local recurrences were observed. Treatment was well tolerated with no grade ≥3 toxicity. Five pts did not have a CR to planned RA therapy including 1 pt that refused additional RT, one pt treated with rituximab, one pt that had a PR on initial evaluation but has not returned for subsequent in person evaluations due to COVID, one pt being observed with stable disease and a final pt that received an additional 20 Gy to the orbit that has a persistent stable mass after the 20 Gy. In a planned subset analysis of 26 pts with newly diagnosed stage 1 disease (MALT, n=22;FL, n=3;low grade BCL, n=1);92.3% (n=24) had a CR to RA RT, with one pt requiring an additional 20 Gy. For all 26 pts with newly diagnosed stage 1 disease, the 3-year freedom from distant relapse rate was 90.4% with 3 distant relapses (contralateral orbit, n=2;paratracheal nodes, n=1). The median follow-up among the 55 pts (MALT, n=38;FL, n=13;low grade B-cell lymphoma, n=4) treated in the retrospective cohort between March 2013 and October 2021 was 28.7 months (95% CI 21.2 - 36.1);98% (n=54) of pts had a CR with RA RT, including 2 pts with a CR after an additional 20 Gy. The remaining pt went on to receive systemic therapy in lieu of additional RT for persistent disease. Among the 54 pts that had a CR with RA RT there was one local relapse in a pt with conjunctival FL 27.8 months after experiencing a CR to ULD RT. This pt received 20 Gy with resolution of the locally relapsed disease. We observed excellent disease control with negligible toxicity in the first prospective study assessing this novel approach of RA ULD RT for pts with indolent B-cell lymphoma. [ FROM AUTHOR] Copyright of International Journal of Radiation Oncology, Biology, Physics is the property of Pergamon Press - An Imprint of Elsevier Science and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full . (Copyright applies to all s.)
ABSTRACT
The pandemic of coronavirus disease 2019 (COVID-19) is rapidly spreading all over the world. In order to reduce the workload of doctors, chest X-ray (CXR) and chest computed tomography (CT) scans are playing a major role in the detection and following-up of COVID-19 symptoms. Artificial intelligence (AI) technology based on machine learning and deep learning has significantly upgraded recently medical image screening tools, therefore, medical specialists can make clinical decisions more efficiently on COVID-19 infection cases, providing the best protection to patients as soon as possible. This paper tries to cover the latest progress of automated medical imaging diagnosis techniques involved with COVID-19, including image acquisition, segmentation, diagnosis, and follow-up. This paper focuses on the combination of X-ray, CT scan with AI, especially the deep-learning technique, all of which can be widely used in the frontline hospitals to fight against COVID-19. © 2022, ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering.
ABSTRACT
The 3-chymotrypsin-like protease (3CLpro) is an attractive target for the development of antiSARS (severe acute respiratory syndrome) drugs. In this work, a series of oxazolidinone derivatives 3a-3v were synthesized and their inhibitory activities against SARS coronavirus 2 (SARS-CoV-2) 3CLpro were evaluated by the fluorescence resonance energy transfer (FRET)-based enzymatic assay. Among synthesized compounds, 3g displayed the best inhibitory activity, with a half maximal inhibitory concentration (IC50) value of 14.47 mu M. Also, docking studies implied that compound 3g was fitted into the active pocket of 3CLpro, forming a hydrogen bond with Glu166.
ABSTRACT
The effects of temperature and relative humidity on the growth of coronavirus disease 2019 (COVID-19) remain unclear. Data on the COVID-19 epidemic that were analyzed in this study were obtained from the official websites of the National Health Commission of China and the Health Commissions of 31 provinces in China. From January 26 to February 25, 2020, the cumulative number of confirmed COVID-19 cases in each region was counted daily using data from our database. Curve fitting of daily scatter plots of the relationship between epidemic growth rate (GR) with average temperature (AT) and average relative humidity (ARH) was conducted using the loess method. The heterogeneity across days and provinces was calculated to assess the necessity of using a longitudinal model. Fixed-effect models with polynomial terms were developed to quantify the relationship between variations in the GR and AT or ARH. An increased AT markedly reduced the GR when the AT was lower than -5°C, the GR was moderately reduced when the AT ranged from -5°C to 15°C, and the GR increased when the AT exceeded 15°C. ARH increased the GR when it was less than 72% and reduced the GR when it exceeded 72%. The temperature and relative humidity curves were not linearly associated with the GR of COVID-19. The GR was moderately reduced when the AT ranged from -5°C to 15°C. When the AT was lower or higher than -5°C to 15°C, the GR of COVID-19 increased. An increased ARH increased the GR when the ARH was lower than 72% and reduced the GR when the ARH exceeded 72%.
ABSTRACT
Although online teaching has been encouraged for many years, the COVID-19 pandemic has promoted it on a large scale. During the COVID-19 pandemic, students at all levels (college, secondary school, and elementary school) were unable to attend school. To maintain student learning, most schools have adopted online teaching. Therefore, the purpose of this study was to explore the design of online teaching activities and online teaching processes adopted by teachers at all levels during the pandemic. Online questionnaires were administered to teachers in Taiwan who had conducted online teaching (including during the formal suspension of classes or simulation exercises) due to the pandemic. According to a quantitative analysis and lag sequential analysis, the instructional behaviors most frequently performed by teachers were roll calls, lectures with a presentation screen, in-class task (assignment) allocation, and whole-class synchronous video-/audio-based discussion. Thus, there were six common significant sequential behaviors among teachers at all levels that were categorized into the four instructional stages of identifying the teaching environment, teaching the class, discussing and evaluating learning effectiveness. College teachers reminded students of some matters first and then called the roll after the students went online. Secondary school teachers were more likely to arrange practical or experimental courses and to use synchronous and asynchronous interactive activities. Finally, elementary school teachers were more likely to use homemade videos and share their screens for teaching and to arrange a large variety of teaching interactions. The differences among colleges, secondary schools, and elementary schools were identified, and suggestions were made accordingly.
ABSTRACT
Background: In this study, we estimated the predictive factors of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission in anesthesiologists performing endotracheal intubation in patients with confirmed SARS-CoV-2. Method: We analyzed data from a survey conducted by the Chinese Society of Anesthesiology Task Force on Airway Management on endotracheal intubation in 98 patients with SARS-CoV-2 confirmed through nucleic acid testing and chest computed tomography. The multivariate logistic model with stepwise selection was used for selecting the predictive factors significantly associated with SARS-CoV-2 infection in the corresponding anesthesiologists. Results: SARS-CoV-2 prevalence in the corresponding anesthesiologists was 20.41% after intubation in patients with SARS-CoV-2. Univariate analysis indicated that intubation for elective treatment, intubation in an operating room or isolation ward, and routine rapid induction with continuous positive-pressure ventilation (PPV) for intubation were associated with a lower SARS-CoV-2 risk in the anesthesiologists. Multivariate analysis revealed that intubation for elective treatment was associated with a significantly decreased SARS-CoV-2 risk (adjusted odds ratio [aOR] = 0.28, 95% confidence interval [CI]: 0.14-0.68, P < 0.0001), and coughing by patients during endotracheal intubation was associated with a significantly increased SARS-CoV-2 risk (aOR = 1.70, 95% CI: 1.39-2.97, P = 0.0404) in the anesthesiologists. Conclusion: Endotracheal intubation for elective treatments, intubation in an operating room or isolation ward, and routine rapid induction with continuous PPV for patients with confirmed SARS-CoV-2 are associated with a lower risk of SARS-CoV-2 transmission in practicing anesthesiologists, and coughing by patients during intubation increases the risk.