Advanced bioimpedance analysis for infectious disease risk assessment via neural network classifiers.

In this work, a neural network classification model based on multidimensional bioimpedance measurement to analyze biomaterial impedance in living systems was developed. The modified Voigt model was used to capture the structural elements as a bioimpedance model. Utilizing this model, extracted descriptors were used to train neural network classifiers. A multidimensional probing technique was employed to obtain biomaterial descriptors, and then Coles plots were generated. Iterative algorithms were applied to generate Voigt models that represent the biomaterial impedance. The model parameters were then utilized as descriptors for the trained classifiers. The developed hybrid classifiers employing these bioimpedance models and descriptor generation algorithms demonstrated their effectiveness in assessing the risk of infectious diseases and related complications. Sample of patients with viral pneumonia were investigated experimentally to evaluate the performance of the bioimpedance models. Bioimpedance analysis was performed by attaching an electrode belt to the patients' chests, and the results were used to generate Cole plots. This innovative approach in bioimpedance analysis has the potential to revolutionize the diagnosis and treatment of infectious diseases. By leveraging advanced technology and algorithms, we can improve the accuracy of infection risk assessment and mitigate its potential complications. This not only enhances patient outcomes but also aids in reducing the transmission of infectious diseases. Furthermore, a comparative analysis was conducted on a control sample of positive and negative cases of pneumonia using X-ray and bioimpedance methods. The bioimpedance method demonstrated an accuracy of 79%, surpassing the X-ray method by 77%.