Towards a Diagnostic Tool for Diagnosing Joint Pathologies: Supervised Learning of Acoustic Emission Signals

Abstract

Acoustic emission (AE) testing detects the onset and progression of mechanical flaws. AE as a diagnostic tool is gaining traction for providing a tribological assessment of human joints and orthopaedic implants. There is potential for using AE as a tool for diagnosing joint pathologies such as osteoarthritis and implant failure, but the signal analysis must differentiate between wear mechanisms—a challenging problem! In this study, we use supervised learning to classify AE signals from adhesive and abrasive wear under controlled joint conditions. Uncorrelated AE features were derived using principal component analysis and classified using three methods, logistic regression, k-nearest neighbours (KNN), and back propagation (BP) neural network. The BP network performed best, with a classification accuracy of 98%, representing an exciting development for the clustering and supervised classification of AE signals as a bio-tribological diagnostic tool.

Acoustic emission signal filtering based on artificial intelligence

“…Logistic regression, Decision tree, Support vector machine (SVM), and K-nearest neighbour (KNN) methods are used in this study for this purpose. The most common method of machine learning used in predicting binary classification problems is logistic regression, in which sigmoid function is used to map real numbers from 0-1 using a given set of features or independent variables [14]. Decision tree classifier is another type of classification-based algorithm, which sorts the classes from root to leaf and terminal nodes.…”

On the Early and Affordable Diagnosis of Joint Pathologies Using Acoustic Emissions, Deep Learning Decompositions and Prediction Machines

“…Tribological research work by Olorunlambe et al [10,11,15] identified the notion that there mainly exists a constrained array of signal processing techniques applied towards ortho-tribological case studies, which they built upon by developing machine learning models that could differentiate between adhesive and abrasive wear conditions. The work by Olorunlambe et al [10,11,15] showcased how machine learning-based methods could distinguish between these two wear conditions from benchtop-based wear conditions in their running in early and steady-state (continuous) stages of wear.…”

“…Tribological research work by Olorunlambe et al [10,11,15] identified the notion that there mainly exists a constrained array of signal processing techniques applied towards ortho-tribological case studies, which they built upon by developing machine learning models that could differentiate between adhesive and abrasive wear conditions. The work by Olorunlambe et al [10,11,15] showcased how machine learning-based methods could distinguish between these two wear conditions from benchtop-based wear conditions in their running in early and steady-state (continuous) stages of wear. The research presented in this paper aims to build on this by designing pattern recognition-based predictive models that can identify and differentiate between the latent and early stages of joint-wear pathologies, a broad class of common joint-wear conditions that includes abrasive, adhesive, burnishing, burnishing-to-scoring transition, and scoring.…”

“…The concept behind AE sensing is based on the use of a passive monitoring technique capable of recording source events that emanate from events of various kinds during energy release [15]. The said energy release is known to travel in the form of stress waves in high-frequency regions (unless stated as audible), which are detected and measured by a sensor that converts an equivalent stress magnitude into a corresponding electrical signal [15]. A visual illustration is shown in Figure 2, while an illustration of the acoustic spectrum can be seen in Figure 3.…”

“…The data collection sequence employed in this paper follows the sequence described by Olorunlambe et al [15], where AE signals were acquired in a setup comprising a tribometer and an AE signal acquisition platform, which is designed to simulate the various wear conditions described in the prior section. An image of the experimental layout can be seen in Figure 7 below.…”