RESUMEN
Generative adversarial networks (GANs) have been recently applied to medical imaging on different modalities (MRI, CT, X-ray, etc). However there are not many applications on ultrasound modality as a data augmentation technique applied to downstream classification tasks. This study aims to explore and evaluate the generation of synthetic ultrasound fetal brain images via GANs and apply them to improve fetal brain ultrasound plane classification. State of the art GANs stylegan2-ada were applied to fetal brain image generation and GAN-based data augmentation classifiers were compared with baseline classifiers. Our experimental results show that using data generated by both GANs and classical augmentation strategies allows for increasing the accuracy and area under the curve score.
Asunto(s)
Procesamiento de Imagen Asistido por Computador , Redes Neurales de la Computación , Encéfalo/diagnóstico por imagen , Imagen por Resonancia Magnética , UltrasonografíaRESUMEN
It is unclear why COVID-19 ranges from asymptomatic to severe. When SARS-CoV-2 is detected, interferon (IFN) response is activated. When it is insufficient or delayed, it might lead to overproduction of cytokines and severe COVID-19. The aim was to compare cytokine and IFN patterns in children and adults with differing severity with SARS-CoV-2.It was a prospective, observational study, including 84 patients. Patients with moderate/severe disease had higher cytokines' values than patients with mild disease (p< 0.001).Two IFN genes were selected to build a decision tree for severity classification: SOCS1 (representative of the rest of the IFN genes) and CIITA (inverse correlation). Low values of CIITA and high values of SOCS1 indicated severe disease. This method correctly classified 33/38(86.8%) of children and 27/34 (79.4%) of adults. To conclude, patients with severe disease had an elevated cytokine pattern, which correlated with the IFN response, with low CIITA and high SOCS1 values.