Acoustic Emission as a non-invasive biomarker for quantitative dynamic assessment of knee joint repair in clinical trials and stratified medicine

Knee osteoarthritis remains a common cause of pain and disability worldwide because there have been very few advances in the treatment of this condition. However, increased research in recent years has started to produce several promising new treatments. These might either improve the condition or reduce the rate at which deterioration occurs.

In order to decide whether new treatments really work, and if so which patients they work best for, it is important to have better techniques to measure whether the joint is getting better or worse. Such techniques should be safe, convenient, and widely available to doctors and other health professionals working in general practices or in hospitals.

Over the past few years we have been researching a new and very distinctive technique for measuring the severity of knee osteoarthritis. This is done by capturing and analysing high-frequency sounds coming from moving knee joints to find out how severely the joint is affected by age and osteoarthritis, and to assess whether the joint is improving or deteriorating.

We have put together a team of experts from the NHS, universities and industry, with a wide range of expertise in medicine, health, science and engineering, to work together on this project. We are using an advanced sound wave sensing technology that is highly sensitive to sound inaudible to the human ear. The equipment was developed originally by engineers for testing and checking for faults in moving parts and in complex structures, such as aircraft.

We attach a small microphone-like sensor to the skin over the knee and collect high-frequency sounds which are produced during sit-to-stand movements. We also attach a small flexible sensor to measure the knee angle, which enables us to know exactly where in the movement cycle each 'sound' is generated. We analyse the amount of 'sound' produced, and also the features of the many different 'sound' waveforms. Our results so far show very clear differences between the 'sounds' from knees of people with and without knee osteoarthritis, and suggest that 'joint sounds' change with age. These very interesting results suggest that it will eventually be possible to use this technology in clinical practice.

We want now to find out whether this technique could be used to provide a better and more convenient way of testing whether or not new treatments work, and of identifying which patients get the most benefit. This work could potentially revolutionise clinical trials in knee osteoarthritis by providing a safe, convenient, portable tool which is widely-available in hospitals and general practices. This would make it possible to do large clinical trials, involving many centres across the world, which would provide better and faster ways of testing urgently-needed new treatments for this condition.

We are the first to investigate high-frequency acoustic emission (AE) as a non-invasive, quantitative biomarker of the integrity of interactions between knee joint components during weight bearing movement (1-4). Our portable system uses wide-band AE sensors to detect sound waves with frequencies up to 200 kHz emitted from knees during sit-stand-sit movement. Similar principles are utilised in non-destructive testing and condition monitoring of engineering structures for early detection of damage and material defects. We are investigating various AE applications in clinical practice and research, and are progressing this work towards prototype development (see Project Plan).
In this study we shall undertake technical validation of AE as a biomarker for enhancing the ability of clinical trials to provide relevant, robust evaluation of the growing array of new interventions aimed at improving or reducing deterioration in joint condition in knee OA. This study will enable us to investigate AE's utility to provide definition of patient subgroups for inclusion in clinical trials, and to inform stratified medicine approaches for optimising the use of such treatments in clinical practice.
AE sensors are piezoelectric transducers which are attached to knee surfaces, using defined anatomical landmarks, to record short bursts of acoustic energy generated by stress upon, and friction between, joint components during weight-bearing movement. An electrogoniometer is attached to enable each acquired AE waveform to be linked to knee angle. A range of signal processing, analysis and visualisation methods has been developed based on AE waveform features in both time and frequency domains, AE statistics in different joint movement phases, and AE classification by principal component analysis. Our published results demonstrate that AE can distinguish not only between healthy and osteoarthritic knees, but also between knees in different age groups and differences in joint condition

The study will address four core impact objectives, which will be met both during and beyond the project's lifetime:

1. Establishing AE biomarkers as distinctive new tools for clinical trials in knee OA
Establishing AE biomarkers for use in clinical trials in knee OA would have a major impact on primary and secondary healthcare as well as benefit for the biopharmaceutical and health technology sectors. There is currently a paucity of biomarkers for knee OA, and all current techniques have important limitations in terms either of their objectivity, their utility and/or their convenience.
The use of simple, non-invasive, portable and easily-accessible equipment, providing an objective measurement, would enable more frequent and comprehensive patient assessment, better definition of patient subgroups, and better targeting and monitoring of treatments. Furthermore, the need for some other assessments, particularly X-rays, should be reduced, thus conserving resources and minimising radiation exposure.
Availability of AE biomarkers would also stimulate increased interest in developing new treatments and interventions for knee OA, and in undertaking large multicentre clinical trials to evaluate their efficacy, whilst also opening up new approaches for improving longer term outcomes. In principle, this may lead to substantial health and economic benefits, not only in terms of improved musculoskeletal health of the population, but also in terms of improved health and performance of the workforce and reduced costs of medical and surgical healthcare.

2. Stimulating development within the health and commercial sectors of new treatments and interventions for knee OA, and consequently increased clinical trials activity in this field.
If the results of this study support the principle, benefits and utility of AE for use in clinical trials, the next stage would be to design and develop prototype equipment tailored for use by clinical researchers and research practitioners, together with associated software, for wider use in future studies. We envisage that it would be feasible to reach such a stage within two years of completion of this project.

3. Achieving effective interdisciplinary, cross-sectoral working
This project depends and draws upon a unique mix of skills and expertise within the project team, which spans musculoskeletal science, medical imaging, signal processing and biostatistics, and upon successful cross-sectoral working. As such, it provides an excellent environment for staff to acquire valuable and stimulating experience of interdisciplinary, cross-sectoral research collaboration, and opportunities for professional development in this context.

4. Stimulating interest in the health and medical devices sector in exploring AE applications for other musculoskeletal and medical conditions.
This work will not only consolidate our leading position in AE applications for knee OA but also provide a foundation for investigating AE in other joint conditions and areas of medicine. By stimulating the development of new devices for acoustic applications in health it offers potential for the UK to take an internationally-leading role in this field.