Clinical and Outcome Study Of COVID-19 At Semi Urban Medical College Hospital in Telangana

Corona virus infection rapidly spreading and producing morbidity and mortality in all over the world over the past one and half year. The virus triggered immune system dysfunction leading to life threatening cytokine storm indicating severe forms of lung injury .Need to understand the clinical profile and risk factors leading to mortality is much needed . Aim(s): To determine the clinical profile of patients having COVID 19 using the inflammatory markers at a semi urban center Methods and materials: A retrospective study conducted on cases that were admitted in CAIMS during the period of 3 months, with CT chest grading CORADS > 3, COVID RT-PCR or rapid antigen test positive, pulse oximetre saturation less than 90% Conclusion(s): 515 cases were taken into study,clinical presentation was observed .Most cases were likely to have CT chest CORADS grading > 3, inflammatory markers like LDH, Sr Ferritin, CRP have been elevated. Cases have shown high IL-6, which was estimated selectively in cases with oxygen support suggesting cytokine storm.16.4 % cases showed mortality. This is attributed among cases with severe form of Covid 19.Copyright © 2020 Ubiquity Press. All rights reserved.

Detection and Diagnosis of Covid-19 from Chest X-Ray Images

The Novel Coronavirus 2019 (COVID-2019) spread quickly around the planet and turned into an undermining pandemic. The early detection of Covid disease is one of the principal challenges and needs on the planet. Early detection helps in controlling the spread of infection. Deep learning has acquired great significance in clinical picture investigation, illustrating improved performance compared to conventional machine learning framework. In this work, a DL based model is proposed for recognizing and characterizing the inconsistencies in chest X-Ray pictures and arranging as unaffected, Covid affected, or Pneumonia. Considering the information inadequacy in clinical space, VGG architecture is utilized as the pre-trained models for building the model for recognition. The quality of the X-Ray pictures and noise present in the pictures influence the decision making leading to high false positives and false negatives. In the current model, pre-processed pictures are fed as input to the DL model accomplishing a maximum accuracy of 96.56%. The proposed model outperforms the DL model without pre-processing with a false positive rate of 0.024 and a false negative rate of 0.026. © 2021 IEEE.