Aussie company, with roots in Perth, has developed an Artificial Intelligence-based mobile application that can accurately assess a person’s pain using a snapshot of their face.
Expressing your pain may sound like a simple enough task, but for people affected by advanced dementia, which impairs their communication skills, pain can be a silent suffering. With this problem in mind, Prof Jeff Hughes and his team from Curtin University, has created a digital approach to help health professional identify patients in pain.
PainChek® is a smart phone-based application that can assess and monitor pain, based on a person’s face. The app seeks to help people with moderate to severe dementia, who have lost the ability to verbally communicate, as well as people who have experienced a stroke, or suffer from a disability that prevents them from communicating with others about their pain.
The idea for creating PainChek® came to Prof Hughes while doing medication management reviews for aged care facilities, where he did not see a lot of change in people’s pain medication.
“I was thinking to myself, well, you know, if I was an elderly patient, in an aged care facility and I’ve got some underlying pain-related condition, then you would imagine over time that there might be a need for a change in pain medication,” Prof Hughes said. “So, I looked at that from two perspectives, firstly/ were there residents who no longer needed to be on pain medications as they may be adversely affecting their quality of life. Take, for example, an elderly patient who’s taking, paracetamol tablets four times a day where they are being crushing up so that they can swallow them, which can leave a severe bitter taste in the mouth, when in fact they no longer need them all the time,” he added.
Secondly, Prof Hughes said, “There’s the situation that a person’s pain is getting worst over time, which can trigger behavioural changes. In many of these cases the pain was not been recognised as the cause, and the patients are treated inappropriately for the behavioural problem, not for the underlying pain”.
According to Prof Hughes, PainChek® is a valuable tool in certain settings, like aged care facilities, where recognising a patient’s pain can lead to changes in their treatment. “It’s estimated in nursing homes, up to 85 per cent of residents have chronic pain. So, they have chronic pain. And with that, they will also have other things that can contribute to acute pain. This might include problems with ulcers in the mouth or teeth abscesses or constipation or a whole range of other things,” Prof Hughes says.
With PainChek® at hand, a healthcare professional or carer can simply use their phone to make an informed assessment about the resident’s pain status.
A WA case example
Back in October 2018, PainChek® was introduced at Regents Garden, an aged care facility in Scarborough, WA, hosting over 500 aged residents. “We have been using PainChek® for monitoring pain in residents with Dementia and doing pain assessments for residents who are non-verbal or have any impairments in communication. The outcome has been very positive through these years,” says Alina Zhang, Specialist Support Manager at Regents Garden.
“Nursing staff stated that the tool has helped the staff to notify or assess pain easily for residents with advanced dementia or in palliative care. According to the high scores, they approach to GPs, specialists or Palliative care team to provide support for the pain management. The process has helped to improve the quality of life for those residents. They also reported that PainChek® has helped residents who suffer from pain in the legs / knees to improve their mobility, when GP’s ordered analgesia’s for them due to the moderate or severe pain results,” she added.
Currently, over 800 aged care facilities are using PainChek®, giving more than 50,000 people living with dementia or cognitive impairment access to this valuable app.
How does it work?
The app uses an Artificial Intelligence (AI), applied to face recognition and analysis to determine if a person is likely to be in pain. “So, it looks at 72 points on the face that represent the structure of the face. But the key thing is that it looks for the brow, the eyes, the nose and the mouth. And then it uses a series of algorithms, nine in total, to detect nine action units, which are muscle contractions or relaxations which indicate the presence of pain,” Prof Hughes explains. “The user also documents non-facial pain behaviours which the residents is exhibiting using a series digital checklist in the app, to allow calculation of a pain score and assignment of pain intensity.”
The whole process may sound complicated, but it only takes 1-2 minutes to do a PainChek® pain assessment. Another advantage of this app is that, rather than relying only on a single person’s face, PainChek® uses an independent assessment. It relies on a massive amount of population data, weaved into its algorithms, which allows it to determine if a face is in pain, based on the identification of specific muscle action units that are indicative of pain.
The road ahead
Despite the clear benefits of using this app, not everyone is on board, Prof Hughes says, with some people still being reluctant at trying the new technology. Prof Hughes expects this will change over time as people become more comfortable with using technology as part of patient care, and more people adopt the app in practice. In the future, Prof Hughes says that a new app has been developed, again using AI, and validated for assessing pain in infants.