Deep learning representations to support COVID-19 diagnosis on CT slices. / Representaciones basadas en aprendizaje profundo como apoyo del diagnóstico de la COVID-19 en cortes de tomografía computarizada

INTRODUCTION: The coronavirus disease 2019 (COVID-19) has become a significant public health problem worldwide. In this context, CT-scan automatic analysis has emerged as a COVID-19 complementary diagnosis tool allowing for radiological finding characterization, patient categorization, and disease follow-up. However, this analysis depends on the radiologist's expertise, which may result in subjective evaluations. OBJECTIVE: To explore deep learning representations, trained from thoracic CT-slices, to automatically distinguish COVID-19 disease from control samples. MATERIALS AND METHODS: Two datasets were used: SARS-CoV-2 CT Scan (Set-1) and FOSCAL clinic's dataset (Set-2). The deep representations took advantage of supervised learning models previously trained on the natural image domain, which were adjusted following a transfer learning scheme. The deep classification was carried out: (a) via an end-to-end deep learning approach and (b) via random forest and support vector machine classifiers by feeding the deep representation embedding vectors into these classifiers. RESULTS: The end-to-end classification achieved an average accuracy of 92.33% (89.70% precision) for Set-1 and 96.99% (96.62% precision) for Set-2. The deep feature embedding with a support vector machine achieved an average accuracy of 91.40% (95.77% precision) and 96.00% (94.74% precision) for Set-1 and Set-2, respectively. CONCLUSION: Deep representations have achieved outstanding performance in the identification of COVID-19 cases on CT scans demonstrating good characterization of the COVID-19 radiological patterns. These representations could potentially support the COVID-19 diagnosis in clinical settings.

Introducción. La enfermedad por coronavirus (COVID-19) es actualmente el principal problema de salud pública en el mundo. En este contexto, el análisis automático de tomografías computarizadas (TC) surge como una herramienta diagnóstica complementaria que permite caracterizar hallazgos radiológicos, y categorizar y hacer el seguimiento de pacientes con COVID-19. Sin embargo, este análisis depende de la experiencia de los radiólogos, por lo que las valoraciones pueden ser subjetivas. Objetivo. Explorar representaciones de aprendizaje profundo entrenadas con cortes de TC torácica para diferenciar automáticamente entre los casos de COVID-19 y personas no infectadas. Materiales y métodos. Se usaron dos conjuntos de datos de TC: de SARS-CoV-2 CT (conjunto 1) y de la clínica FOSCAL (conjunto 2). Los modelos de aprendizaje supervisados y previamente entrenados en imágenes naturales, se ajustaron usando aprendizaje por transferencia. La clasificación se llevó a cabo mediante aprendizaje de extremo a extremo y clasificadores tales como los árboles de decisiones y las máquinas de soporte vectorial, alimentados por la representación profunda previamente aprendida. Resultados. El enfoque de extremo a extremo alcanzó una exactitud promedio de 92,33 % (89,70 % de precisión) para el conjunto 1 y de 96,99 % (96,62 % de precisión) para el conjunto-2. La máquina de soporte vectorial alcanzó una exactitud promedio de 91,40 % (precisión del 95,77 %) para el conjunto-1 y del 96,00 % (precisión del 94,74 %) para el conjunto 2. Conclusión. Las representaciones profundas lograron resultados sobresalientes al caracterizar patrones radiológicos usados en la detección de casos de COVID-19 a partir de estudios de TC y demostraron ser una potencial herramienta de apoyo del diagnóstico.

Atypical ST-Segment-Elevation Myocardial Infarction Presentation in Patients With COVID-19 at a High-Volume Center in New York City.

Atypical presentations of ST-segment-elevation myocardial infarction (STEMI) have been reported in patients who have COVID-19. We have seen this occurrence in our center in Bronx, New York, where multitudes of patients sought treatment for the coronavirus. We studied the prevalence of atypical STEMI findings among patients with COVID-19 who presented during the first 2 months of the pandemic. Consistent with previous reports, 4 of our 10 patients with COVID-19 and STEMI had no identifiable culprit coronary lesion; rather, they often had diffuse ST-segment elevations on surface electrocardiograms along with higher levels of D-dimer and inflammatory markers. In contrast, 32 of 33 patients without COVID-19 (97%) had a culprit lesion. The patients with COVID-19 and a culprit lesion more often needed thrombectomy catheterization and administration of glycoprotein IIb/IIIa inhibitors. Our study confirms that patients with COVID-19 often have atypical STEMI presentations, including the frequent absence of a culprit coronary lesion. Our findings can help clinicians prepare for these atypical clinical presentations.

[Early Detection of COVID-19 in Portugal: Use of Clinical Records]. / Deteção Precoce de COVID-19 em Portugal: Uso de Registos Clínicos.

INTRODUCTION: Syndromic surveillance allows early detection of changes in the population's morbidity pattern. The aim of this study is to evaluate the usefulness of indicators related to access to healthcare services, in COVID-19 surveillance. MATERIAL AND METHODS: A time series analysis was performed using the weekly incidence rate of COVID-19 in Mainland Portugal, between weeks 14/2020 (March 30 to April 5) and 25/2020 (June 15 to 21), and six indicators: 1) COVID-19 consultations in primary healthcare; 2) number of COVID-19 emergency department visits; 3) number of emergency department visits due to viral pneumonia; 4) number of hospitalizations due to viral pneumonia; 5) proportion of emergency department visits due to viral pneumonia; and 6) proportion of hospitalizations for viral pneumonia. Pearson correlation and cross-correlations were computed. RESULTS: A strong correlation was found between the weekly incidence rate of COVID-19 and all indicators. [(1) 0.76; (2) 0.82; (3) 0.77; (4) 0.84; (5) 0.86; e (6) 0.90]. Emergency department visits and hospitalizations for viral pneumonia detect variations in the frequency of the disease with a one week lag compared to the incidence rate of COVID-19, in one week. COVID-19 consultations in primary healthcare and emergency department visits trail behind the incidence rate of COVID-19, in one week. The proportion of viral pneumonias in emergency department visits, or hospitalizations, is temporally aligned with the weekly incidence rate of COVID-19. DISCUSSION: The delay found in the COVID-19 primary healthcare consultations and emergency department visits, may be related to changes in access to healthcare services and clinical coding. Emergency department visits and hospitalizations for viral pneumonia may be useful in the early detection of COVID-19. Viral pneumonia may have been coded as being of unknown origin. Future monitoring of these indicators is necessary to ascertain whether the incidence of COVID-19 is significantly influenced by changes in testing strategies. The indicators described in this study will be an asset for the optimization of testing strategies, allocation of healthcare resources to the communities that are most vulnerable to severe morbidity and assessing vaccination impact. As such, surveillance systems based on clinical data will be a valuable complementary tool to SINAVE. CONCLUSION: The indicators under analysis could be used regularly, with special attention to viral pneumonias, to detect outbreaks of COVID-19. Information on pneumonia of unknown etiology may be considered in the surveillance of COVID-19.

Introdução: A vigilância sindrómica permite a identificação precoce de alterações no padrão de morbilidade da população. Este estudo tem como objetivo avaliar a utilidade de indicadores relativos a cuidados de saúde primários e hospitalares, na vigilância da COVID-19.Material e Métodos: Foi realizada uma análise de séries temporais utilizando a taxa de incidência semanal de COVID-19 em Portugal Continental, entre as semanas 14/2020 (30 março a 05 abril) e 25/2020 (15 a 21 junho), e seis indicadores: 1) consultas em cuidados de saúde primários por COVID-19; 2) número de episódios de urgência por COVID-19; 3) número de episódios de urgência por pneumonia vírica; 4) número de internamentos por pneumonia vírica; 5) proporção de episódios de urgência por pneumonia vírica face ao total de episódios de urgência por pneumonia; e 6) proporção de internamentos por pneumonia vírica face ao total de internamentos por pneumonia. Foram calculadas correlações de Pearson e correlações cruzadas.Resultados: Foi encontrada uma correlação forte entre a taxa de incidência semanal de COVID-19 e todos os indicadores [(1) 0,76; (2) 0,82; (3) 0,77; (4) 0,84; (5) 0,86; e (6) 0,90]. Os episódios de urgência e internamento por pneumonias víricas detetam variações na frequência da doença, com uma semana de antecedência. As consultas em cuidados de saúde primários e urgências por COVID-19 registam uma semana de atraso relativamente à evolução da taxa de incidência. A proporção de pneumonias víricas face ao número de pneumonias em episódios de urgência, ou internamentos, encontra-se alinhada temporalmente com a evolução da taxa de incidência semanal de COVID-19.Discussão: O atraso encontrado no padrão de evolução de consultas em CSP, e de episódios de urgência por COVID-19 face à incidência de COVID-19, poderá estar relacionado com a reorganização dos serviços de saúde e criação de códigos específicos para estas consultas. Episódios de urgência e internamentos por pneumonia vírica poderão ser úteis para a deteção precoce de possíveis surtos de COVID-19. Pneumonias víricas poderão ter sido classificadas como pneumonias de causa indeterminada. A monitorização futura destes indicadores é necessária de modo a averiguar se a incidência de COVID-19 é influenciada significativamente por alterações na estratégia de testagem. Os indicadores deste trabalho serão uma mais valia para a adequação de estratégias de testagem, alocação de recursos de saúde a comunidades mais vulneráveis à morbilidade severa e avaliação de programas de vacinação. Como tal, os sistemas de vigilância com base em registos de saúde serão um complemento valioso ao SINAVE.Conclusão: Sugere-se que os indicadores em análise sejam utilizados de forma regular, com especial atenção à informação relativa a pneumonias víricas, como forma de detetar precocemente surtos de COVID-19. A informação relativa a pneumonias de causa indeterminada poderá ser considerada na monitorização da COVID-19.

Preprocedural Pool Testing Strategy for Dentistry during the COVID-19 Pandemic.

INTRODUCTION: Aerosol-generating procedures (AGPs) put the dental health care professionals (DHCPs) at a greater risk for acquiring severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. In late June 2020, the Centers for Disease Control and Prevention advised elective dental procedures provision to asymptomatic patients while mandating strict infection control protocol and suggested the use of preprocedural testing as an adjunct. A cost-effective method for mass preprocedural testing is pool testing, which has specificity and sensitivity similar to polymerase chain reaction. This article aims to assess the outcomes and utility of incorporating preprocedural testing protocol for SARS-CoV-2 in dental clinics before providing AGPs. METHOD: The patients who were recommended AGPs where rubber dam placement was not possible were advised to undergo preprocedural testing for SARS-CoV-2. Pool testing strategy was employed, and patients were asked to get tested 48 h before the day of the procedure. RESULTS: Out of a total of 1,000 patients, who presented from June 2020 to late July 2020, 464 were recommended dental procedures. In 194 of 464, AGPs could not be performed under rubber dam isolation; therefore, the patients were advised to get a preprocedural pool test. In total, 111 patients deferred the procedure and testing. Out of 83 who got tested, 7 were positive for SARS-CoV-2, 5 of whom were tested in early June 2020 and 2 in late July 2020. CONCLUSION: Pool testing within its limitations can be a useful preprocedure test in asymptomatic low-risk patients for AGP in dentistry, especially when the disease prevalence is low or moderate (<10%). It has the potential of reducing testing costs significantly while conserving reagent and other resources. Preprocedure testing, however, also gives rise to certain ethical concerns that also need to be addressed. KNOWLEDGE TRANSFER STATEMENT: The results of this study can be used by clinicians when deciding which preprocedure testing approach they wish to use when performing aerosol-generating procedures in asymptomatic patients with consideration of cost sensitivity and specificity values.

A comparison of swab types on sample adequacy, suspects comfort and provider preference in COVID-19.

AIM: This study was aimed to compare the virological, suspect reported outcomes and provider preferences during COVID-19 swab taking procedure used for sampling. METHODS: The COVID-19 suspects are subjected to nasopharyngeal (NP) and oropharyngeal (OP) swabs for testing. Two types of swabs (Nylon and Dacron) are used for sample collection. Prospectively each suspect's response is collected and assessed for self-reported comfort level. The provider's experience with each suspect and virological outcomes recorded separately. The sample adequacy was compared based on swab types and demographic characteristics. RESULTS: A total of 1008 COVID-19 suspects were considered for comparison of various outcomes. Dacron and flocked Nylon swab sticks are used for taking 530 and 478 samples, respectively. Suspects who underwent the procedure using Nylon swabs were six times more likely to have pain/discomfort compared to when Dacron swab was used (Adj RR (95% CI: 6.76 (3.53 to 13, p=0.0001))). The providers perceived six times more resistance with the Nylon swabs compared to Dacron Swabs (Adj RR (95% CI: 5.96 (3.88 to 9.14, p=0.0001))). The pediatric population had a higher rate of blood staining in Dacron swab [Dacron 66 (80.5%); Nylon 51 (54.8%) p=0.0001]. The sample adequacy rate and laboratory positivity rate were not significantly different from each other. CONCLUSIONS: Given the comparable virological outcomes, the difference in suspect and providers comfort should drive swab selection based on characteristics of the suspects. The bulbous Nylon swab caused more pain/discomfort in adults compared to Dacron.