Adipose Tissue Characterization With Electrical Impedance Spectroscopy and Machine Learning

Abstract

Biological tissues have variable passive electrical properties depending on their cellular constitution. Electrical impedance spectroscopy (EIS) is commonly used to monitor cell and tissue characteristics. By measuring the impedance of a sample at various frequencies, it is possible to collect information regarding cell size and shape, cell membrane properties, or cytoplasm conductivity. From the perspective of longitudinal structural monitoring, bioimpedance measurements outrank traditional tissue analysis methods, such as fixation and slicing, owing to their nondestructive nature. Machine learning can be used to automatically process the impedance data and make real-time classifications of tissue types. Here, we present preliminary results on ex-vivo mouse adipose tissue measurements using EIS and further data processing and classification using machine learning models.