Project Summaries - Cohort 1
In its first cohort, the South Yorkshire Digital Health Hub allocated nearly £500,000 to seven innovative projects that will positively impact healthcare in the region. These projects are:
Mobile technology to enhance prehabilitation care for joint replacement patients
Lead investigator: Dr Andrew Barnes, Sheffield Hallam University
a.barnes@shu.ac.uk
What is the problem you are aiming to solve?
Joint replacement is common to reduce the physical limitations of osteoarthritis. Prehabilitation improves postoperative outcomes, but there is a need to optimise these care pathways.
What is your overall aim?
To develop a mobile system with a focus on resistance training for tracking and providing feedback to support the pre-rehabilitation phase of joint replacement patients.
Vision
The long term vision of this project is to develop a commercial cost-effective mobile system which supports prehabilitation care pathways. A sample of ten participants comprising six people with lived experience of osteoarthritis and four service providers from the NHS or third sector will be recruited to participate in the co-design and evaluation of the prototype technology. The proposed devices of interest are either wearable sensors or camera-based smartphone applications which capture data on performance of resistance exercise.
Objectives
Determine user preference for either wearable or camera based technology.
Develop a mobile system for tracking and providing feedback to support the pre- rehabilitation phase of joint replacement patients.
Conduct a small scale pilot study with joint replacement patients (proof of concept).
DMO4LNC - Digital mobility outcomes for long term neurological conditions
Lead Investigator: Dr. Ellen Buckley, University of Sheffield
e.e.buckley@sheffield.ac.uk
What is the problem you are aiming to solve?
There is a lack of tools for accurate assessment and monitoring of walking in real-world settings in many long term neurological conditions.
What is your overall aim?
To develop a digital technology approach to study real-world walking for early identification and monitoring in long term neurological conditions.
Vision
We want to develop a remote digital monitoring tool to capture real-world gait and mobility that can provide vital information to aid early diagnosis and monitor disease progression. This information could aid clinical decision making and selection of appropriate interventions. A scalable digital technology tool for remote mobility monitoring could also have use for characterising and monitoring the real-world mobility of people with a range of long term neurological conditions.
Objectives
Month 1-3: Stakeholders workshops to explore acceptability issues around approaches to monitor mobility remotely
Month 4-8: User pilot testing in cohort of adults with Cerebral Palsy, Progressive Supranuclear Palsy or Normal Pressure Hydrocephalus including in-clinic characterisation and 14day remote assessment
Month 5-10: Qualitative work with pilot participants regarding acceptability and feasibility of implementation within clinical pathways and identification of potential barriers.
Month 6-12: Develop data analysis pipelines for processing of digital outcome measures for disease stratification including comparison with an existing dataset
How do we make diagnosis of coronary artery disease faster, more convenient, more equitable and more cost effective?
Lead investigator: Rafic Ramses, University of Sheffield
rramses1@sheffield.ac.uk
What is the problem you are aiming to solve?
Coronary artery disease is the UK’s biggest killer and yet its diagnosis is slow, inconvenient, inequitable, and expensive.
What is your overall aim?
To discover if Artificial Intelligence applied to smartphone, wearable, and routinely collected NHS data can predict underlying coronary artery disease.
Vision
Our vision is to develop community-based diagnostics for CAD. This could make diagnosis faster, reduce the need for expensive CT scans, and reduce healthcare inequalities.
Chest pain due to coronary artery disease is usually triggered by physical activity and increased heart rate. Our partner Checkpoint Cardio has developed a wearable sensor that measures continuous ECG, heart rate and blood pressure at home for days or months. Smartphones and wearables are also able to measure physical activity and heart rate. This data, if collected during episodes of chest pain, may indicate whether someone has CAD. However, data such as medical history also affects risk of CAD and this can be found in the NHS record.
Our approach tests whether applying artificial intelligence to Checkpoint Cardio, smartphone, wearable, and NHS data can predict underlying coronary artery disease.
Objectives
Design and build an integrated set of devices that record physical activity, heart rate, chest pain episodes, ECG, & blood pressure.
Recruit 63 patients with and without CAD
Collect continuous heart rate, ECG, blood pressure, activity data and timing of episodes of chest pain from participants
Build Machine Learning models that uses the sensor, demographic, and NHS data to predict coronary artery disease
The Hormone Effect (T.H.E) project - Co-design of a digital data collection tool to capture lived experience of hormonal contraception
Lead investigator: Dr Rebecca Mawson, University of Sheffield
r.l.mawson@sheffield.ac.uk
What is the problem you are aiming to solve?
The lack of real-world data on hormonal contraception effects leads to distrust, misinformation and hesitancy. Black and ethnically diverse women are worse affected by this, especially those living in poverty.
What is your overall aim?
To create a self-reported data collection tool to capture lived experiences of hormonal contraception. Designed by women who are under-represented in mainstream research.
Vision
At its core, this project is about inclusivity and women's health. By using digital technology to explore the lived experience of under-represented communities, we aim to address significant gaps in reproductive health research. We strive to create a health inequalities exemplar project with under-represented and underserved populations as the driving force.
Using a self-reported data collection tool would be the first time we will have explored the lived experience of hormonal contraception in real-life settings, opening up new avenues for understanding and improving women's health. We envision scaling up to become the hormonal version of Zoe Health Study, with the usability of the period tracker Flo app, thereby making a tangible and transformative difference for women in the UK and worldwide.
Objectives
Build collaboration with women from under-represented communities to explore digital health technology use via public engagement and PPIE.
Bridge the divide in digital accessibility using pre-existing inclusive health research networks in Sheffield DeepEnd Research Alliance (DERA).
Rebuild trust in the use of technology through fully co designed dynamic methods with women under-represented in mainstream research.
Capture lived experience of women taking hormonal contraception via a longitudinal pilot in general practice in areas of highest unmet reproductive health need.
The Hormone Effect Project website
MIRACLE - Medical Imaging Reports Accessible and CLear for Everyone
Lead investigator: Dr Samer Alabed, University of Sheffield
s.alabed@sheffield.ac.uk
What is the problem you are aiming to solve?
40 million medical imaging scans are performed in the UK yearly. 96% of their reports don’t make sense to patients, leading to anxiety and burdening mental health. Clinicians spend valuable clinic time explaining results instead of discussing management.
What is your overall aim?
Real-time simplification of medical imaging reports, enhancing readability and patient understanding without compromising medical accuracy.
Vision and approach
The success of this project is when patients read their medical imaging report, nod, and understand what it says clearly. Simplifying medical imaging reports is the Renaissance of healthcare communication, akin to the historical shift of translating Latin texts—once exclusive to clergy and scholars—into the language of the lay people. This is not a distant dream —we have it within our grasp with the expertise, datasets and technology available to us.
This SYDHH project will continue to lead to a Pilot Product with iterative refinement and validation to create report summaries for multiple end-users: patients, clinicians, multi-disciplinary meetings, discharge summaries. The technology will be expanded to simplify reports of all imaging modalities (MRI, CT, ultrasound and X-ray) in various diseases. Industry partners will be sought to make the tool widely available.
Objectives
Technology development. Develop a reliable system to simplify medical imaging reports, making them accessible and understandable for patients without compromising accuracy.
Evaluate the effectiveness of simplified reports on patient comprehension and engagement.
Assess the impact on patient satisfaction, healthcare efficiency and patient outcomes.
Ensure scalability and sustainability of the solution in diverse patients and various imaging modalities and diseases.
ExAMPLe - Explainable AI-based Multi-modal Prediction in Lung cancer treatment
Lead Investigator: Dr. Bilal Tahir, University of Sheffield
b.tahir@sheffield.ac.uk
What is the problem you are aiming to solve?
Previous lung cancer survival prediction and treatment response assessments relied on mono-modal data streams, failing to capture the complex interplay of factors influencing patient outcomes.
What is your overall aim?
Develop an innovative, explainable multi-modal AI model integrating radiological imaging, histopathology, and clinical tabular data for enhanced lung cancer survival prediction and treatment response assessment.
Vision
Our long-term vision is to transform lung cancer care through personalised, AI-driven survival prediction and treatment response assessment. This project will lay the groundwork by developing a prototype explainable AI model that integrates radiological imaging, histopathology, and clinical data, demonstrating the potential of multi-modal AI in lung cancer prognosis and paving the way for future clinical translation and value creation in the digital health sector. By harnessing the resources and expertise of the SYDHH, this pump-priming study will lay the foundation for future larger-scale projects and clinical translation, ultimately contributing to the development of personalised cancer care and the transformation of the digital health economy.
Objectives
Curate diverse data types (radiological, histopathological, clinical) to create a robust pipeline for multi-modal AI analysis by month 4.
Develop and preliminarily validate a prototype explainable AI model integrating diverse data types by month 8.
Actively engage with patients and public throughout to ensure clinical relevance, trust, and alignment with real-world needs.
Lay foundation for future larger-scale projects and clinical translation, exploring value creation in digital health by month 12.
Wearable tech to detect bleeding in patients receiving anti clotting drugs
Lead investigator: Dr William Parker
w.parker@sheffield.ac.uk
What is the problem you are aiming to solve?
Anti-clotting drugs reduce cardiovascular risk but increase bleeding. Bleeding may start out as silent before causing harm. It is difficult to conveniently detect silent bleeding.
What is your overall aim?
Use wearable digital health technology to enhance the monitoring of patients receiving anti-clotting drugs, detecting bleeding earlier and intercepting problems before they cause major harm.
Vision
Our vision is to improve the safety of anti-clotting treatment to prevent harm and save lives. Major bleeding is common in patients taking anti-clotting medications, particularly when more than one type is given together. We want to detect early, unnoticed signs of bleeding such as a drop in blood count, to prevent progression to the most harmful, sometimes fatal, events. It means we could investigate and treat the cause of bleeding sooner to avoid serious harm, hospital admission or disability. Our approach must have convenience for patients at its core. We think wearable technology can best achieve this vision and have identified a clear and feasible approach for testing. Our proposed work will take us towards this vision by proving and refining the concept, testing it in the population we are trying to help and identifying any further barriers to overcome.
Objectives
Perfect the measurement of blood count using wearable technology.
Test if wearable technology can be used to measure changes in blood count in the target population and compare to gold-standard blood tests.
Assess the benefits and barriers of the proposed solution for patients.
Ensure the proposed solution works across people of different genders, ethnicities and ages.