Lung perfusion findings on perfusion SPECT/CT imaging in non-hospitalized de-isolated patients diagnosed with mild COVID-19 infection

Autor: Osayande Evbuomwan, Gerrit Engelbrecht, Melissa V. Bergman, Sello Mokwena, Oluwatosin A. Ayeni
Jazyk: angličtina
Rok vydání: 2021
Předmět:
Zdroj: The Egyptian Journal of Radiology and Nuclear Medicine, Vol 52, Iss 1, Pp 1-12 (2021)
Druh dokumentu: article
ISSN: 2090-4762
DOI: 10.1186/s43055-021-00521-1
Popis: Abstract Background The aim of this retrospective study is to assess the incidence and type of lung perfusion abnormalities in non-hospitalized patients diagnosed with mild COVID-19 infection after de-isolation. Data from 56 non-hospitalized patients diagnosed with COVID-19 infection referred to our nuclear medicine department from July–December 2020 for a perfusion only SPECT/CT study or a ventilation perfusion SPECT/CT study were collected. Images were assessed for the presence and type of perfusion defects. The CT component of the study was also assessed for the presence of mosaic attenuation and COVID pneumonia changes. Results Thirty-two (57.1%) cases had perfusion defects. There were 20 (35.7%) cases with defects in keeping with pulmonary embolism, 17 (30.4%) cases with defects associated with mosaic attenuation but not due to pulmonary embolism, and 6 (10.7%) of cases with defects due to pulmonary infiltrates from COVID pneumonia. A total of 24 (42.9%) cases had mosaic attenuation on CT, with 10 (17.9%) of them showing a pattern likely consistent with shunting on the perfusion images. Conclusion Lung perfusion abnormalities are a common finding in non-hospitalized COVID-19 patients with mild disease. They are usually either due to pulmonary embolism, parenchymal infiltrates, or other causes of mosaic attenuation related to, but not specific to the pathophysiology of COVID-19 infection. The value of VQ SPECT/CT imaging is also shown in this study, in detecting and differentiating the various types of perfusion abnormalities.
Databáze: Directory of Open Access Journals