A new class of phenomena identified during the analysis of neural network of multidimensional data from patients with inflammatory lung diseases

Aim: to apply the neural network analysis method of multi-data of patients with community-acquired pneumonia (CAP) to recognize the variability of their diagnoses and discovery of new analytical capabilities of NSA based on new methodological approaches of the meeting. Materials and methods: in this paper, we applied a new approach of neural network analysis of multivariate data, obtained based on clinical, laboratory and instrumental tests in 60 patients older than 65 years at various stages of the treatment of community-acquired pneumonia. Anthropometric data were used together with the results of immunological and immunochemical studies of blood serum of patients with community-acquired pneumonia who were in the acute phase of the disease. Results: a new approach analysis of these data revealed the presence of characteristic groups of the bio-markers, which consist from a combination of a small number of the signs that are necessary and sufficient in the aggregate for accurate classification of diagnoses for patients. Conclusions: the presence of these symptoms characteristic series shows that we have discovered a new class of phenomena. These phenomena manifest themselves in the hidden relationships between the signs which included in these groups and are reflect features of the flow processes in the pathogenesis of inflammatory diseases of the lungs in different diagnostic areas. Their study can be an important and interesting in terms of understanding the many aspects of this disease.