2021, issue 3, p. 34-42
Received 14.06.2021; Revised 30.06.2021; Accepted 28.09.2021
Published 30.09.2021; First Online 25.10.2021
https://doi.org/10.34229/2707-451X.21.3.3
Previous | Full text (in Ukrainian) | Next
Bayesian Recognition Procedures with Independent Signs of Inflammatory Processes in Gliomas, Metastases and Meningiomas by Indicators of Erythrocyte Sedimentation Rate
Andrii Tarasov
V.M. Glushkov Institute of Cybernetics of the NAS of Ukraine, Kyiv
Correspondence: This email address is being protected from spambots. You need JavaScript enabled to view it.
Introduction. The article discusses the application of Bayesian recognition procedures with independent signs in relation to the data of the modified erythrocyte sedimentation rate, which were taken from patients with gliomas, metastases, meningiomas, craniocerebral concussion and from a group of healthy people.
Purpose of the article. Improving the efficiency of recognition of inflammatory processes in gliomas, metastases and meningiomas by indicators of erythrocyte sedimentation rate using optimal recognition procedures with independent signs.
Results. In previous articles by the authors, an attempt was made to recognize inflammatory processes by indicators of the modified erythrocyte sedimentation rate caused by brain cancer using Bayesian recognition procedures based on a single substance. In this work, a new model was built using several independent signs (different substances) at once. The results obtained on the basis of the new model significantly increased their efficiency in relation to the models that were used earlier. Such an increase in all comparisons ranged from 3 to 12 %, and up to almost 94 %. If earlier it was possible to recognize only combinations of diagnoses in which there were no more than two diagnoses, then in this work for the first time it was possible to recognize three diagnoses at once. At the same time, the recognition efficiency became slightly more than 70 %. An attempt was also made to recognize more than three diagnoses, but the new model did not give significant results, slightly exceeding 50 % when recognizing four diagnoses at once.
Conclusions. Thanks to the use of Bayesian recognition procedures with independent signs, it was possible to significantly increase the recognition of inflammatory processes caused by brain cancer. The modified erythrocyte sedimentation rate, which is an auxiliary tool in the diagnosis of gliomas, allows one or another pathology to be determined in the preoperative period, since the pathology is finally determined only when studying a surgically removed tumor. In the postoperative period, such a modification is an indicator of repeated recurrence of gliomas. It was also possible to significantly increase the recognition of inflammatory processes caused by non-oncological disease (traumatic brain injury) in relation to oncological processes in gliomas, metastases and meningiomas.
Keywords: Bayesian recognition procedure, independent signs, gliomas, metastases, meningiomas, modified erythrocyte sedimentation rate, complex parameter.
Cite as: Tarasov A. Bayesian Recognition Procedures with Independent Signs of Inflammatory Processes in Gliomas, Metastases and Meningiomas by Indicators of Erythrocyte Sedimentation Rate. Cybernetics and Computer Technologies. 2021. 3. P. 34–42. (in Ukrainian) https://doi.org/10.34229/2707-451X.21.3.3
References
1. Gridina N.Ya., Sergienko I.V. Analysis of Speed Indexes of Erythrocyte Sedimentation for Cerebral Gliomas. Journal of Automation and Information Sciences. 2007. 39 (12). P. 52–59. https://doi.org/10.1615/JAutomatInfScien.v39.i12.50
2. Tarasov A.L., Gupal A.M., Gridina N.Ya. Modofication of the Use of Bayesian Recognition Procedures for Inflammatory Processes in Gliomas, Metastasis and Meningiomas by Indicators of Erythrocyte Sedimentation Rate. Cybernetics and Computer Technologies. 2021. 2. P. 57–62. https://doi.org/10.34229/2707-451X.21.2.5
3. Gridina N.Ja., Gupal A.M., Tarasov A.L. Bayesian Recognition of Inflammatory Processes in Brain Gliomas. Cybernetics and Systems Analysis. 2017. 53 (3). P. 366–372. https://doi.org/10.1007/s10559-017-9936-4
ISSN 2707-451X (Online)
ISSN 2707-4501 (Print)
Previous | Full text (in Ukrainian) | Next