How Facial Recognition Works

Pain is manifested in a change of physical, emotional and behavioural traits. Facial expressions are the most specific encodable form of pain behaviour in humans, as well as being accessible and highly plastic and dynamic.1

The facial expression of pain has emerged as an important pain indicator in people with dementia who have difficulties in verbalising their pain.2


People with dementia lose the ability to suppress their learnt pain behaviours, and express their pain through facial expressions more frequently and intensely. The PainChek app has been validated in clinical trials to understand pain intensity of this population.

PainChek® App Features

Automated facial recognition analysis:

•Three second video of patient face recognises 9 micro-facial expressions that indicate the presence of pain

Digital questionnaire checklist:

•guide the carer in other pain assessment factors e.g. movement, vocalization leading questions with Yes/No decisions


Automated pain assessment score:

•based on 42 test points


Documented electronically via cloud backend:

•pain trend line and monitoring of treatment
•capability to integrate into electronic medical records

How does PainChek® Work?

PainChek is a secure smartphone or tablet based medical device that uses artificial intelligence and your input, to assess a person’s level of pain. PainChek provides a pain score in real time, which will help you gauge whether treatment is required. This information is then securely stored in the cloud and if you chose to, can be shared with your GP or healthcare professional.

PainChek allows you, the carer to identify and measure a person’s level of pain and supports you to assess and manage a person’s pain experience. PainChek calculates a pain score and intensity (ie No Pain, Mild, Moderate or Severe). This then allows you, the carer to intervene if pain is present and manage pain accordingly.

It is important to note that PainChek will not give you advise on what to do if pain is detected. In these instances you should always seek medical advice from your healthcare professional.

It is also important to assess and reassess the person’s pain after every time you have tried to manage their pain, to ensure the process of pain management is working. In any circumstances when you are unsure or have concerns about the well-being of the person you are caring for, please contact your healthcare professional.

Automated reporting features enable you to see times and patterns of a person’s pain.

The Technology Behind PainChek®

PainChek uses Artificial Intelligence (AI) based computer vision algorithms called “convolutional and deep neural networks” to detect the facial expression features and characteristics most consistent with acute and chronic pain. The networks were trained through a mathematical process of machine learning.  Images of people in pain were analysed  by experts and the pain markers identified and coded. The coded data was then used to train the PainChek algorithm to automatically identify facial pain indicators in the most accurate way.

The pain detection algorithm is integrated into the PainChek app and uses a smartphone or tablet device’s camera to briefly observe a person’s face. This video observation is analysed in real time using facial detection software. All of this occurs locally on the device and no images are captured, stored or transmitted. The facial analysis is then combined with other indicators of pain, such as vocalisations, behaviours  and movements to calculate a pain intensity score.


With continuous development and improved hardware capabilities, PainChek’s AI will become the world’s leading expert in pain facial expression analysis and pain behavioural patterns.

•1 A. Williams. Facial expression of pain: An evolutionary account. Behavioural and Brain Sciences(2002) 25, 439–488.
•2.M Kunz et al (2008). The facial expression of pain in patients with dementia. Pain. 133. 221-8. 10.1016/j.pain.2007.09.007.