A smartphone app that uses facial recognition to automatically detect pain in non-communicative people with dementia could revolutionise care and dramatically improve quality of life.
The Electronic Pain Assessment Tool (ePAT) being developed by researchers at Curtin University uses the mobile phone’s camera and facial recognition software to automate the facial recognition component of pain assessment.
The results are combined with the information collected on the other accepted pain indicators – vocalisation, behavioural change, psychological change, physiological change and physical change — to provide a total pain score.
The tool is currently being tested on an Android smartphone platform but it will ultimately be available for Android and Apple phones and tablets.
By automating the facial recognition component, the tool aims to provide an objective way to detect an accurate level of pain in non-communicative people with dementia, said the tool’s lead developer, Professor Jeff Hughes from Curtin’s School of Pharmacy.
“We think it could be an absolute game changer and set the gold standard in that there should be an objective way in which to determine the presence of pain in people who have communication difficultly,” Prof Hughes said.
“The tool is extremely quick. The results that you get will be reproducible. You will be able to collect the pattern of the patient’s pain over time.”
Prof Hughes said the tool will first be able to detect and quantify pain, after which it can be used to re-score the patient to determine the effectiveness of the analgesic given.
How it works
The app records a five-to-10 second video of the person’s face using the smartphone’s camera. Looking at nine features of the face it locates markers to determine which of the particular indicators of pain are there, Prof Hughes explained.
The video is taken, utilised but not stored, and as everything is done in real time it is not a privacy risk, he said.
The tool then takes the user through the remaining six domains of pain indication where they input the information using the phone’s touch screen.
“At the end it gives you what it estimates as the final score and then it gives you a severity scale,” Prof Hughes said. “Once you get familiar with it, it would take less a minute to do the whole consultation.”
If we can guarantee whether a patient is in pain or not and can prove the objectivity of using ePAT we can ensure that everyone who has pain will be treated, he said.
Better pain management will improve a person’s behaviour and cognition, making them easier to be cared for, he said. “It will have an enormous impact on the quality of life for patients with dementia and importantly it will also have an impact on the quality of life of carers.”
As the tool is “extremely” easy to use, it will be suitable, and empowering, for family carers as well professional carers and health professionals, he said.
An exciting project
Director of HammondCare’s Dementia Centre Colm Cunningham agreed this tool could have a positive impact on the quality of life of someone with dementia.
Associate Professor Cunningham is currently preparing a paper for HammondCare’s upcoming international dementia conference about pain management in people with dementia and the under recognition of pain.
“I am promoting the human facial assessment of pain through facial recognition with those markers you are talking about it. From the sounds of it they are being more accurate because we often keep it to simpler facial recognition markers so the complexity I think is very exciting,” Prof Cunningham said.
Prof Cunningham said it looked like the tool was aiming to advance what was already known – that facial recognition was a really good way of assessing people.
“The level of untreated pain in people with dementia is very high … so having this tool is a thumbs up from my point of view,” he said.
Prof Cunningham said the Dementia Centre would be focusing on the culture of care in terms of pain management for people with dementia in a new three-year research project called INTERVENE.
He said while there were processes for pain management – he was taught assess, plan, implement, evaluate when he trained as a nurse in 1986 – the processes were not always applied.
“If the tool reinforces the process of good pain management then it has achieved a massive amount because that is the issue, people do not return and assess and see if they have had successful pain relief.”
Testing the tool
The first validation phase is planned for February when the researchers will test the tool with older people without cognitive impairment.
The team is then aiming to begin a residential aged care facility-based pilot before June, which will test the tool with residents with cognitive impairment and compare the usability and the validity of the package against the standard Abbey tool.
“Once we have done our validation, we can then work on the facial features, not only to detect but also to quantify the pain. There is data from the literature to say that this can be done,” Prof Hughes said.