Abstract
Background. Prediction of cross-sectional area n. vagus on the sonogram, based on the values of some inflammation biomarkers in the blood
Methods. In 68 patients with a PCR test (+) for SARS-COV 2, the cross-sectional area of the nerve (CSA) was measured by ultrasound, as an average of three measurements. The values of CRP, fibrinogen and ferritin were determined. CSA values from both sides and all biomarkers were classified as (0)- within normal limits and (1)- all other values. Logistic linear regression was performed and a strong linear relationship between predictors was excluded. A neural network and binary logistic regression model was formed for the CSA values on the right side, on the left side the connection between the selected predictors and the CSA of the left vagus was not established.
Results. On the right side, CSA is 1.32±0.3; on the left 1.2±0.28. In the study, 11.8%, 22.1% and 8.82% had normal values of CRP, fibrinogen and ferritin respectively. A strong linear association of predictors was ruled out by linear logistic regression. A neural network with two hidden layers and 6 number of units in either hidden layer was formed in 2 seconds. The accuracy percentage of the model in training is 91.5% and in testing it is 90.5%. The AUC of this model is 0.76 (p=0.018, p<0.05). Logistic binary regression model: logitX2D= 18.018+ (20.752x CRP values)+ (-38.77x fibrinogen values)+(2.175x ferritin values), where X2D- is the cross-sectional area of the right n. vagus, given by the model. Exp(B) of the model is 7.5, Psample = 0.88, OR=56.25, accuracy of the model is 91.2%, Omnibus test: χ2=14.193, p=0.003(p<0.05). Model reliability coefficient -2LogLH=35.068 and AUC =0.76, p=0.018. Overall model quality 0.61 (limit is 0.5). On the left side, the value of CSA n.vagus cannot be predicted using selected predictors.
Keywords
Array
Array
Array
Array
Array
Array
Array
References
The published articles will be distributed under the Creative Commons Attribution 4.0 International License (CC BY). It is allowed to copy and redistribute the material in any medium or format, and remix, transform, and build upon it for any purpose, even commercially, as long as appropriate credit is given to the original author(s), a link to the license is provided and it is indicated if changes were made. Users are required to provide full bibliographic description of the original publication (authors, article title, journal title, volume, issue, pages), as well as its DOI code. In electronic publishing, users are also required to link the content with both the original article published in Journal of Medical Biochemistry and the licence used.
Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.