Tissue classification from electric impedance spectroscopy for haptic feedback in minimally invasive surgery

Abstract

Haptic feedback is missing in teleoperated surgical robots creating a sensory disconnect from the surgeon and their patient. This thesis proposes using the electric impedance of tissues, instead of the traditionally used mechanical impedance, to develop haptic feedback for surgical robots. Electric impedance spectroscopy (EIS) and a modified surgical needle were successfully able to measure the electric impedance for gel-based phantoms, ex-vivo tissue, and freshly excised organs. Processes for fitting the electric impedance of these tissues to the double-dispersion Cole model were developed including stochastic and deterministic approaches. The tissues were classified with least square error, k-Nearest Neighbour and Naïve Bayes using the measured electric impedance and the extracted model parameter values. The thesis culminates in applications of using EIS as part of implementing vibrotactile and force feedback applications involving sets of user trials to validate its effectiveness in identifying the tissue through haptic feedback.