Determination of blood dielectric properties based on frequency and blood gas indicators

Abstract Blood is a vital component of the human body, and alterations in its dielectric properties can result in changes to the dielectric properties of vital organs and even the entire body. This, in turn, affects the accuracy of bioelectromagnetic techniques such as thoracic electrical bioimpedance, electrical impedance tomography (EIT) and the body composition analysis and body fluid management based on biological impedance measurement. To detect brain damage signals by EIT during cardiopulmonary bypass, where significant alterations in blood composition are observed, it is first necessary to investigate the change of blood dielectric parameters. Accordingly, the purpose of this study is to elucidate the blood dielectric parameters of patients through blood gas indicators and frequency. A correlation analysis was used to ascertain the blood gas indicators that exert a significant influence on the blood dielectric parameters within 10 Hz–100 MHz. Based on these findings, a series of regression models were constructed to determine the dielectric parameters. This study is the first to demonstrate that pH, partial pressure of oxygen (PO 2 ), partial pressure of carbon dioxide (PCO 2 ), K + , and Ca 2+ concentrations significantly influence blood dielectric properties. The results of the cross-validation and statistical analysis demonstrated that the Gaussian process regression is an effective method for accurately determining the blood dielectric parameters based on blood gas indicators and frequency, with good generalization performance. This finding has the potential to facilitate the effective extraction of brain damage signals by EIT, while also advancing the field of biophysics, biosensors, bioelectromagnetic technology, and electromagnetic modeling of the human body.