Breakthrough detection for orthopedic bone drilling via power spectral density estimation of acoustic emission

Abstract

Breakthrough detection of robotic drilling process for orthopedic surgery is proposed. Features are extracted from Power Spectral Density (PSD) with Welch methods from acoustic emission signal during the drilling operation. According to the power spectrum, Band-Pass Infinitive Impulse Filter is performed to efficient frequency band selection. Most meaningful features which have been derived as median frequency and mean frequency are selected to create a projected feature. A simple breakthrough algorithm, which compares the current value of the projected feature with an adaptive threshold, is developed.