Computer Vision to Record Catheter Output
Research type
Research Study
Full title
Computer Vision to Automatically Record Urinary Catheter Output: An Exploratory Study
IRAS ID
261066
Contact name
Stefan Williams
Contact email
Sponsor organisation
Leeds Teaching Hospitals NHS Trust
Duration of Study in the UK
2 years, 0 months, 1 days
Research summary
A significant proportion of hospital inpatients have a urinary catheter – a tube from the bladder that empties into a measuring bag next to the hospital bed. The measuring bag is called a ‘urometer’. It is crucial to measure the level of urine in the urometer at regular intervals, to detect and prevent kidney damage. At present, nursing staff check the urometer bag at intervals and write down the volume of urine each time. Unfortunately, at least half of the time, this documentation is incomplete, and serious kidney damage can go undetected until it has become irreversible.
Our exploratory study aims to understand whether the level of urometer fluid could be automatically detected using computer processing of simple camera images (as a potential automatic measuring system). To do this, we intend to collect a large number of photographs of urometer bags in real-world conditions (varying lighting, colour, angles etc), and use these photographs to test computing methods for their ability to detect the urometer fluid level. Anonymous photographs of urometer bags will be taken by staff in Leeds Teaching Hospitals NHS Trust, who have signed written instructions, including the requirement that absolutely no part of the patient or their belongings be photographed. Patients will be given written and verbal information about the study, including information that they can decline the urometer photography at any time. Urometer photographs will not be taken if patients are acutely unwell or have any potential impairment in the capacity to understand information about the study or communicate the choice to opt out.
We aim to continue this study until we have a large number of photographs of urometer catheter bags, between 1000 and 3000 photographs, to provide a range of images sufficient for the testing of computer techniques to detect the urometer fluid level.
REC name
North East - York Research Ethics Committee
REC reference
19/NE/0140
Date of REC Opinion
23 Apr 2019
REC opinion
Favourable Opinion