Electrical Impedance Spectroscopy for Tissue Classification

This study utilized electrical impedance spectroscopy (EIS) techniques for classifying various types of ex-vivo biological tissues.  To conduct this research, a quantity of five 20x20x20 mm specimens per tissue type was used for collecting impedance data.  This data was measured using a Quadra Impedance Spectroscopy Device, in which impedance was measured over a range of 15 frequencies from 1 kHz to 349 kHz.  Machine learning-based approaches were then used for classifying the tissues based on the impedance data.  The results showed that frequency was negatively correlated with impedance magnitude, whilst having a positive correlation with impedance phase; there was also greater error when measuring the impedance at lower frequencies.  It was also revealed that the most accurate classifier was Ensemble at roughly 99.13%.  Overall, this study proves that EIS can be an effective tool for classifying biological tissues through impedance acquisition and that its applications could aid doctors in differentiating types of tissue for surgery.