Detection of SARS-CoV-2 pneumonia: two case reports

Peripheral HLA-G/ILT-2 immune checkpoint axis in acute and convalescent COVID-19 patients

The immunosuppressive non-classical human leukocyte antigen-G (HLA-G) can elicits pro-viral activities by down-modulating immune responses. We analysed soluble forms of HLA-G, IL-6 and IL-10 as well as on immune effector cell expression of HLA-G and its cognate ILT-2 receptor in peripheral blood obtained from hospitalised and convalescent COVID-19 patients. Compared with convalescents (N = 202), circulating soluble HLA-G levels (total and vesicular-bound molecules) were significantly increased in hospitalised patients (N = 93) irrespective of the disease severity. During COVID-19, IL-6 and IL-10 levels were also elevated. Regarding the immune checkpoint expression of HLA-G/ILT-2 on peripheral immune effector cells, the frequencies of membrane-bound HLA-G on CD3+ and CD14+ cells were almost identical in patients during and post COVID-19, while the frequency of ILT-2 receptor on CD3+ and CD14+ cells was increased during acute infection. A multi-parametric correlation analysis of soluble HLA-G forms with IL-6, IL-10, activation markers CD25 and CD154, HLA-G, and ILT-2 expression on immune cells revealed a strong positive correlation of soluble HLA-G forms with membrane-bound HLA-G molecules on CD3+/CD14+ cells only in convalescents. During COVID-19, only vesicular-bound HLA-G were positively correlated with the activation marker CD25 on T cells. Thus, our data suggest that the elevated levels of soluble HLA-G in COVID-19 are due to increased expression in organ tissues other than circulating immune effector cells. The concomitant increased expression of soluble HLA-G and ILT-2 receptor frequencies supports the concept that the immune checkpoint HLA-G/ILT-2 plays a role in the immune-pathogenesis of COVID-19.

A MobileNet-based CNN model with a novel fine-tuning mechanism for COVID-19 infection detection

COVID-19 is a virus that causes upper respiratory tract and lung infections. The number of cases and deaths increased daily during the pandemic. Once it is vital to diagnose such a disease in a timely manner, the researchers have focused on computer-aided diagnosis systems. Chest X-rays have helped monitor various lung diseases consisting COVID-19. In this study, we proposed a deep transfer learning approach with novel fine-tuning mechanisms to classify COVID-19 from chest X-ray images. We presented one classical and two new fine-tuning mechanisms to increase the model's performance. Two publicly available databases were combined and used for the study, which included 3616 COVID-19 and 1576 normal (healthy) and 4265 pneumonia X-ray images. The models achieved average accuracy rates of 95.62%, 96.10%, and 97.61%, respectively, for 3-class cases with fivefold cross-validation. Numerical results show that the third model reduced 81.92% of the total fine-tuning operations and achieved better results. The proposed approach is quite efficient compared with other state-of-the-art methods of detecting COVID-19.