Purpose Confronting the pandemic of COVID-19 is nowadays one of the most prominent challenges of the human species. A key factor in slowing down the virus propagation is the rapid diagnosis and isolation of infected patients. The standard method for COVID-19 identification, the Reverse transcription polymerase chain reaction method, is time-consuming and in short supply due to the pandemic. Thus, researchers have been looking for alternative screening methods, and deep learning applied to chest X-rays of patients has been showing promising results. Despite their success, the computational cost of these methods remains high, which imposes difficulties to their accessibility and availability. Thus, the main goal of this work is to propose an accurate yet efficient method in terms of memory and processing time for the problem of COVID-19 screening in chest X-rays. Methods To achieve the defined objective, we propose a new family of models based on the EfficientNet family of deep artificial neural networks which are known for their high accuracy and low footprints. We also exploit the underlying taxonomy of the problem with a hierarchical classifier. A dataset of 13,569 X-ray images divided into healthy, non-COVID-19 pneumonia, and COVID-19 patients is used to train the proposed approaches and other 5 competing architectures. We also propose a cross-dataset evaluation with a second dataset to evaluate the method generalization power. Results The results show that the proposed approach was able to produce a high-quality model, with an overall accuracy of 93.9%, COVID-19 sensitivity of 96.8%, and positive prediction of 100% while having from 5 to 30 times fewer parameters than the other tested architectures. Larger and more heterogeneous databases are still needed for validation before claiming that deep learning can assist physicians in the task of detecting COVID-19 in X-ray images, since the cross-dataset evaluation shows that even state-of-the-art models suffer from a lack of generalization power. Conclusions We believe the reported figures represent state-of-the-art results, both in terms of efficiency and effectiveness, for the COVIDx database, a database of 13,800 X-ray images, 183 of which are from patients affected by COVID-19. The current proposal is a promising candidate for embedding in medical equipment or even physicians' mobile phones.

Background: Musculoskeletal injuries (MSK-I) are a serious problem in sports medicine. Modifiable and nonmodifiable factors are associated with susceptibility to these injuries. Thus, the aim of this study was to describe the prevalence of and identify the factors associated with MSK-I, including tendinopathy and joint and muscle injuries, in athletes. Methods: In this cross-sectional observational study, 627 athletes from rugby (n = 225), soccer (n = 172), combat sports (n = 86), handball (n = 82) and water polo (n = 62) were recruited at different sports training centres and competitions. Athlete profiles and the prevalence of MSK-I were assessed using a self-reported questionnaire. Only previous MSK-I with imaging confirmation and/or a positive physical exam by a specialized orthopaedist were considered. The association of the epidemiological, clinical and sports profiles of athletes with MSK-I was evaluated by a logistic regression model. Results: The mean age was 25 ± 6 years, and 60% of the athletes were male. The epidemiological, clinical and sports profiles of the athletes were different for the five sport groups. The MSK-I prevalence among all athletes was 76%, with 55% of MSK-I occurring in a joint, 48% occurring in a muscle and 30% being tendinopathy, and 19% of athletes had three investigated injuries. The MSK-I prevalence and injury locations were significantly different among sport groups. There was a predominance of joint injury in combat sports athletes (77%), muscle injury in handball athletes (67%) and tendinopathy in water polo athletes (52%). Age (≥30 years) was positively associated with joint (OR = 5.2 and 95% CI = 2.6-10.7) and muscle (OR = 4.9 and 95% CI = 2.4-10.1) injuries and tendinopathy (OR = 4.1 and 95% CI = 1.9-9.3). Conclusion: There is a high prevalence of tendinopathy and joint and muscle injuries among rugby, soccer, combat sports, handball and water polo athletes. The analysis of associated factors (epidemiological, clinical and sports profiles) and the presence of MSK-I in athletes suggests an approximately 4-5-fold increased risk for athletes ≥30 years of age. The identification of modifiable and non-modifiable factors can contribute to implementing surveillance programmes for MSK-I prevention.

BackgroundTendinopathy pathogenesis is associated with inflammation. Regulatory T (Treg) cells contribute to early tissue repair through an anti-inflammatory action, with the forkhead box P3 (FOXP3) transcription factor being essential for Treg function, and the FC-receptor-like 3 (FCRL3) possibly negatively regulating Treg function. FCRL3 –169T>C and FOXP3 –2383C>T polymorphisms are located near elements that regulate respective genes expression, thus it was deemed relevant to evaluate these polymorphisms as risk factors for tendinopathy development in athletes.MethodsThis case-control study included 271 volleyball athletes (146 tendinopathy cases and 125 controls) recruited from the Brazilian Volleyball Federation. Genotyping analyses were performed using TaqMan assays, and the association of the polymorphisms with tendinopathy evaluated by multivariate logistic regression.ResultsTendinopathy frequency was 63% patellar, 22% rotator cuff and 15% Achilles tendons respectively. Tendinopathy was more common in men (OR = 2.87; 95% CI = 1.67–4.93). Higher age (OR = 8.75; 95% CI = 4.33–17.69) and more years of volleyball practice (OR = 8.38; 95% CI = 3.56–19.73) were risk factors for tendinopathy. The FCRL3 –169T>C frequency was significantly different between cases and controls. After adjustment for potential confounding factors, the FCRL3 –169C polymorphism was associated with increased tendinopathy risk (OR = 1.44; 95% CI = 1.02–2.04), either considering athletes playing with tendon pain (OR = 1.98; 95% CI = 1.30–3.01) or unable to train due to pain (OR = 1.89; 95% CI = 1.01–3.53). The combined variant genotypes, FCRL3 –169TC or –169CC and FOXP3 –2383CT or –2383TT, were associated with an increased risk of tendinopathy among athletes with tendon pain (OR = 2.24; 95% CI: 1.14–4.40 and OR = 2.60; 95% CI: 1.11–6.10). The combined analysis of FCRL3 –169T>C and FOXP3 –2383C>T suggests a gene-gene interaction in the susceptibility to tendinopathy.ConclusionsFCRL3 –169C allele may increase the risk of developing tendinopathy, and together with knowledge of potential risk factors (age, gender and years playing) could be used to personalize elite athletes’ training or treatment in combination with other approaches, with the aim of minimizing pathology development risk.

No abstract

No abstract