Towards Non-Invasive Liver Health Monitoring: Comprehensive Microwave Dielectric Spectroscopy of Freshly Excised Human Abdominal Tissues

Abstract

Metabolic dysfunction-associated steatotic liver disease ranks among the most prevalent chronic liver conditions worldwide. To reduce its burden, early diagnosis is vital to enable timely medication and rehabilitation. The non-invasive diagnosis of liver health is challenging due to the limitations of existing methods. For this purpose, the design of portable non-invasive electromagnetic sensors requires knowledge of how human liver tissue and other abdominal tissues interact with electromagnetic waves. This necessitates the accurate characterisation of dielectric properties of the liver and adjacent abdominal tissues. Since postmortem changes or prolonged storage significantly change those properties and lead to incorrect interpretation, fresh human abdominal tissues, including skin, fat, muscle, and liver, were obtained at surgery, and their dielectric properties were measured immediately in the microwave frequency range of 0.5 GHz to 15 GHz. An adaptive weighted vector mean optimization algorithm was used to derive the parameters of a second-order Cole-Cole model using the experimental data. Statistical and cluster analyses were performed on the curated database following the derived model. The results showed that hepatic steatosis significantly changed the dielectric properties of the liver $(p < 0.001)$. Moreover, the liver had distinct dielectric properties from the skin, fat, and muscle tissues $(p < 0.05)$. These findings suggest that electromagnetic sensors could be used to assess liver health in a non-invasive way, which could improve liver health outcomes and reduce costs.