Funded by the Engineering and Physical Sciences Research Council (EPSRC), 2021–2023
Principal Investigator: Dr Stephen Preece, University of Salford
University of Plymouth Co-Investigator: Professor Eduardo Miranda
Co-Investigators: Dr Wendy Maltinsky, University of Stirling; Ms Andrea Taylor, Glasgow School of Art
Multimodal biofeedback
A potentially untapped application is the use of digital technology to empower patients to change unhealthy behaviours and facilitate self-management of conditions like DB. A clinical multimodal biofeedback system that conveys integrated muscle function can provide patients with the skills and motivation to optimise muscle patterns and greatly improve the management of DB and other health conditions. This would move away from resource-intensive treatments that currently involve simple breathing exercises and visual observation by a clinician.
Breathing biofeedback techniques have been shown to be effective for improving outcomes in other conditions, such as cystic fibrosis, COPD and hypertension. The use of behavioural change techniques (e.g., building motivation, setting goals and establishing new habit formation) alongside biofeedback has been shown to have positive outcomes and greater adherence in people with severe asthma.
Dysfunctional breathing (DB)
Integrated muscle function is essential for a range of physiological functions, including postural control, movement and breathing. Suboptimal muscle function has been implicated in numerous conditions, including breathing disorders and musculoskeletal pain, as well being linked to anxiety. In addition, poor muscle function impacts movement coordination and so could affect mobility and motivation to take regular exercise.
Dysfunctional breathing – an inefficient pattern of breathing – has a high prevalence in the UK (approximately 10%). Symptoms include breathlessness, persistent coughing and the feeling of not being able to take a deep breath, but is also associated with back pain, digestive problems, asthma and COPD. Currently, clinical assessments and treatments for DB rely on visual assessments and breath-holding tests/questionnaires without validated tools to measure muscle patterns.
The clinical management of DB
A prototype will be created to identify integrated muscle function from sensor measurements using devices such as 3D cameras and wearable sensors. The user will access data from multimodal biofeedback that uses an individual avatar to visualise complex muscle patterns, with the use of sounds to convey subtle changes in muscle function. High-dimension data can be communicated through variations in timbre, volume, pitch and rhythm.
Integrated with a behaviour change intervention and initially supported by a physiotherapist, this will allow those with DB to be able to re-educate their muscles for effective breathing. Ultimately, this will become a fully autonomous system. Using an untapped application of this technology will empower patients to change behaviours, raising their awareness of the physiological processes that are affecting them, facilitating self-management and ultimately reducing time spent in traditional health care delivery.
Real-time visualisation and sound
The project will first seek to develop software that can produce an avatar matched to the body shape of the user that will show geometrical changes in body shape from breathing (e.g., chest/abdominal expansion and shoulder movements). This will be done in consultation with those with DB. An algorithm will then be created to focus on muscle patterns.
Professor Eduardo Miranda will then lead the participatory design of the biofeedback system that will enable real-time visualisation and sound to represent muscle patterns during breathing. The system will respond in real-time to changes in the muscular control of breathing to allow behaviour change. It will include sonification methods and be customised to the user’s body shape as well as ensuring users can customise the visual and auditory components to their liking. This will facilitate conditioned learning through the association of specific sounds with ‘correct’ breathing.
Delivering digitally-enabled care
OptiMuscle can ultimately deliver towards the NHS Long Term Plan to provide digitally-enabled care across the UK. As a fully autonomous system, OptiMuscle would significantly reduce clinical time and strain on resources and vastly improve health outcomes.
In the long-term, it is hoped this project will create a system that can recognise suboptimal muscle patterns across a number of scenarios without clinical input. This could be used for a range of health conditions, including musculoskeletal pain, mobility in older people through to helping people manage anxiety-related disorders and even long Covid.
Professor Eduardo Reck Miranda
A distinguished composer and AI scientist, Professor Miranda's research is prolific. Eduardo has developed musical algorithms that are controlled by electrical signals from the brain to allow severely disabled former musicians to make music, and ‘biocomputers’ which use live organisms as processors and as such, in a performance context, exhibit a behaviour much more compatible with our own than conventional computers.
Also a member of the University’s Centre for Health Technology, Eduardo’s research is opening up possibilities for people with disabilities, supporting music-based activity for palliative care within occupational and music therapy. He heads the Interdisciplinary Centre for Computer Music Research (ICCMC), which is internationally known for its pioneering research and for making a difference in the world.