Is Machine Learning Able to Detect and Classify Failure in Piezoresistive Bone Cement Based on Electrical Signals?

TL;DR

This study demonstrates that machine learning algorithms, especially neural networks, can effectively analyze electrical signals from piezoresistive bone cement to detect, locate, and classify implant failures, potentially improving early diagnosis.

Contribution

The paper introduces a novel integration of machine learning with electrical impedance tomography and smart piezoresistive materials for failure detection in bone implants.

Findings

Neural networks achieved over 91.9% accuracy in location tracking.

Neural networks achieved over 95.5% accuracy in defect location specification.

Neural networks achieved over 98% accuracy in defect classification.

Abstract

At an estimated cost of $8 billion annually in the United States, revision surgeries to total joint replacements represent a substantial financial burden to the health care system. Fixation failures, such as implant loosening, wear, and mechanical instability of the poly(methyl methacrylate) (PMMA) cement, which bonds the implant to the bone, are the main causes of long-term implant failure. Early and accurate diagnosis of cement failure is critical for developing novel therapeutic strategies and reducing the high risk of a misjudged revision. Unfortunately, prevailing imaging modalities, notably plain radiographs, struggle to detect the precursors of implant failure and are often interpreted incorrectly. Our prior work has shown that the modification of PMMA bone cement with low concentrations of conductive fillers makes it piezoresistive and therefore self-sensing such that when combined with a conductivity imaging modality, such as electrical impedance tomography (EIT), it is possible to monitor load transfer across the PMMA using cost-effective, physiologically benign, and real-time electrical measurements. Herein, we expand upon these results by integrating machine learning techniques with EIT. We survey different machine learning algorithms and principal component analysis for application to this problem, including neural networks on voltage readings of an EIT phantom for tracking position of a sample, specifying defect location, and classifying defect types. Our results show advantage of neural network with more than 91.9 %, 95.5 %, and 98 % accuracy in interpreting EIT signals for location tracking, specifying defect location, and defect classification respectively. These preliminary results show that the combination of smart materials, EIT, and machine learning may be a powerful tool for diagnosing the origin and evolution of failure in joint replacement.