About TORUS

Neurological disorders are the single largest cause of disability – in the UK alone there are 150,000 people with Parkinson’s disease, the fastest-growing neurological condition. Parkinson’s disease is incurable, and symptoms worsen over time, severely reducing quality of life and creating heavy burdens on the patient’s family. The cost to the NHS each year is £375M, with families and social services contributing a further £877M (Centre for Health & Social Care Research, 2017). The number of people with Parkinson’s disease in the UK is expected to nearly double by 2040. 

To get a new drug to market, pharmaceutical (pharma) companies need to evidence by a clinical trial whether the drug improves symptoms such as freezing when walking, tremor and the ability to undertake daily tasks such as standing up from sitting or moving between rooms. Currently, to gather this evidence, each patient in the trial must travel to hospital to be observed performing standardised tests by a clinician. However, these (at most) monthly “snapshot” samples of symptoms are a poor representation of the hour-by-hour variation of the patient’s true symptoms. This problem undermines hugely expensive clinical trials to the extent that some large companies have publicly withdrawn from developing new drugs for Parkinson’s disease. 

This programme’s target is therefore the clear and urgent need for accurate measures to assess new disease-modifying (or curing) medicines for Parkinson’s disease. The vision of TORUS is to create the capability to autonomously, continuously and objectively measure symptoms of illness (mobility-related activities of daily living) many times every day during the clinical trial of a new drug, in the patient’s own home and for months at a time. TORUS will achieve this goal by using a wrist-worn wearable integrated synergistically with AI-enabled cameras. The data from the wearable and cameras is fused to give metrics of the quality of mobility-related activities. The programme concludes with a clinical proof of concept.  

Research Programme and Objectives

The TORUS objectives include trust, privacy, data fusion and distributed machine learning, reflecting not only a commitment to advance new technologies with clinical and patient input, but also a commitment to Responsible Research and Innovation. The project will be working with clinicians and patients, including patient populations which are less well represented in research.

Objective 1: Trustworthy, explainable, privacy-enhanced, distributed data fusion pipeline for quantifying quality of movement at home. 

Objective 2: Optimised wearable battery life and data quality via real-time intelligent reconfiguration. 

Objective 3: Validated, autonomous, secure, privacy-preserving and resilient integrated sensor system suitable for the homes of patients (i. Wearable; ii. AI-enabled camera; iii. Machine Learning system). 

Objective 4: Contextual voice-based annotation system, suitable for patients, to continually-enrich models over time. 

Objective 5: Ensuring that decisions taken in TORUS reflect the views and priorities of diverse patients & families. 

Objective 6: Privacy & informed consent at the heart of the patient experience.  

Objective 7: Determine how best to characterise specific mobility-related activities of daily living in a way that reflects the priorities of patients.

Objective 8: Demonstrate TORUS capability in a patient cohort.  

The TORUS programme consists of six interlinked Work Packages, each comprising specific domain expertise and managed by experts in that discipline. Details are below:

Work Packages

Work Package 1

A Wearable for Data Fusion

The team’s aim is to design a unique TORUS wearable as one component of a data-fusion system. Uniquely, it will be context- and mission-aware continuallyreconfigured by the TORUS system to make the most of its battery, data storage, computation and connectivity. 

Members:

Dr George Oikonomou, Dr Silvia Del Din, Dr James Pope, Dr Zahraa Abdallah, Dr Kirsty Scott, Dr Lisa Alcock, Dr Shuhao Dong

 

Work Package 2

Video Sensing for Data Fusion

The team will focus on using video features, computed in the home, to characterise mobility-related activities of patients over periods of many months. This raw image/video data will be processed into useful but privacy-preserving features of movement by an AI-enabled camera.

Members:

Prof Majid Mirmehdi, Dr Amirhossein Dadashzadeh, Dr Jingjing Liu

Work Package 3

Data Fusion and Distributed Machine Learning

This team will lead on the design, creation and implementation of the TORUS distributed machine learning pipeline, including orchestrating the approach to data fusion as a collaboration with WP1 and WP2, optimising the powerful fusion of wearable and video-derived data. 

Members:

Prof Paul Watson, Dr Telmo de Menezes e Silva Filho, Prof Raul Santos-Rodriguez, Dr Zahraa Abdallah, Prof Majid Mirmehdi, Dr Guanxiong Sun, Dr Chloe Winchliffe

Work Package 4

System Integration, Security, Testing and Trial Support 

This team will lead on designing the overall TORUS system architecture; undertaking all the work required to support the integration of the outcomes of individual work packages into a single, autonomous, secure, privacy-preserving and resilient system. This WP will also develop the TORUS system’s remote monitoring and fault diagnostics.

Members:

Dr George Oikonomou, Prof Paul Watson, Dr Em Tonkin

Work Package 5

PPIE, Codesign, Ethics and Acceptability  

This team will lead on co-design for TORUS, collaborating with other WPs, drawing on methods for empirical Public and Patient Involvement and Engagement (PPIE)It will ensure that key aspects of the TORUS system itself are co-designed with those with lived experience of Parkinson’s disease, plus with stakeholders in industry, charities and the TORUS multidisciplinary Co-Is. 

Members:

Prof Abi Durrant, Prof Dave Kirk, Dr Aisling O’Kane, Prof Kenton O’Hara, Dr Alan Whone, Dr Cathy Morgan, Dr Elaine Czech, Dr Lenia Margariti

Work Package 6

Into the Real World 

 This team will lead TORUS in a real-world proof of concept trial of its technology in a patient cohort. This includes co-design of the trial and consent process with patients, ethical approval, recruiting patients, conducting the trial and analysing the data. 

Members:

Prof Lynn Rochester, Dr Alan Whone, Prof Alison Yarnall, Dr Cathy Morgan

For more information on each member go to research team page.

For details on how to get in touch please go to the contact page.