How Is Facial Recognition Aiding Medical Industry

Facebook
Twitter
LinkedIn
Email
Print

 

COMMUNICATED CONTENT – Facial recognition technology is used in various ways. From your phone that can scan your face and unlock the screen to a social media app that can identify you in a photograph with your friends, the ability to scan your face is vital. When you walk through the airport, a check-in system is capable of identifying you as you stare into a camera. Then, there is law enforcement, using the tech to track down criminals.

How does it work? When it generally involves scanning your facial traits in two or three dimensions. It then plots the physical characteristics of a person, including the angles and the distance from the eyes to the nose and more, to discern who you are. The level of accuracy of the tech is surprisingly high and it can be done hands-free and from a socially acceptable distance. Therefore, not only can it be used to secure your phone, but to identify you and validate your identity in a healthcare setting.

But how and why would the medical sector want to use facial recognition technology? Here we explore the different applications.

Minimizing patient errors

There is a group of researchers in Seoul who are looking to create a facial recognition app for hospitals to prevent confusion between patients. While the idea of being unconscious and mistaken for someone else feels unnerving, the fact there is now a solution is comforting. If you are being transferred between caregivers and you are not conscious, the receiving professional can guarantee your identity with the use of facial recognition technology.

There are also facilities where it is important to monitor the coming and going of patients, such as psychiatric facilities. The technology could be used to alert staff of the comings and goings of individuals allowed time out of the hospital. More effectively, the University in South Korea also believes that the technology can be used to monitor moods as well as identity. Therefore, if a patient is showing signs of depression, anxiety or pain, for instance, the carers could be informed.

Doctor diagnosis tool

There are then the applications of such technology that help doctors to diagnose illnesses. For instance, there was a doctor trying to identify a genetic disorder in a child. She scanned the face of the child into an app and it mapped landmarks on the face and used machine learning to match this to rare genetic disorders. The facial recognition technology helped the doctor to work out what was wrong and then accurately map a future care plan.

 

Even where the genetic disorder is common, the markers on the face are so mild that they might not be discerned by the naked eye. Therefore, a scan of the face and machine learning could make this diagnosis much more accurate. A computer mapping landmarks on the face is always going to be more accurate than the naked eye.

 

The app is known as Face2Gene and is becoming much more widely used, especially in pediatrics. One researcher noted how facial analysis technology was capable of recognizing Down Syndrome in non-caucasian races even with the physical differences across races. Another researcher noted how it stopped bias in diagnosis, which sometimes led to mistakes.

 

Detecting pain

While most pain manifests itself to medical staff in the form of screams or moans, there is a problem of quiet pain. Quiet pain occurs where the patient is in a state or an environment where they cannot express pain. For instance, a newborn baby doesn’t always cry when they are in pain. Indeed, there are some neurological conditions that change the behaviors of babies and can hide problems.

 

Consequently, facial recognition technology can be used to assess pain in newborn babies, which can sometimes be a subjective matter for a caregiver. Due to desensitization to the sounds of a babies’ ward, the observer can miss alerts to problems due to observer bias. 

 

A researcher developed a system called Classification of Pain Expression (COPE) that was capable of analyzing the images of baby expressions and alerting when there was a pain. During the research, the analysis of the face was thought to be 90% accurate when follow-up tests were undertaken. This offers the opportunity to continually monitor the face of a newborn and look for crucial behavioral and physiological problems that might otherwise be missed.

 

The same technology could be used with elderly or incapacitated patients. A system in Australia called PainChek is being used for nonverbal patients, such as those struggling with dementia. It is thought it could also be applied in intensive care units and for people under anesthetic. Involuntary movements during operations could alert the surgeon to a patient that is feeling pain.

 

Helping those with Autism

Facial recognition technology is also being used as a means of helping those with an autism spectrum disorder. Stanford University researchers are using Google Glass and a headset to help track facial expressions in others and communicating emotions to the wearer. Consequently, if a child with autism looks up to an adult whose expression reads “frustrated” then the device will coach the child on how to react.

 

The idea of such technology is not to be used throughout life but to act as a kind of therapy to teach children to recognize facial expressions themselves. There has been substantial evidence that the device has improved socialization in young people and so helped them to enjoy social interactions.

 

Before we get too carried away by the possibilities of Facial Recognition Technology, extensive thought has to be given to the privacy and ethical concerns of the technology. HIPPA does provide a framework for protecting patient data and the FRT scan can be anonymized like any other scan made in the hospital. However, the chance for reidentification with facial scans is much higher and so causes significant concerns.

Consequently, while it is a powerful tool, many in the medical sector wish to proceed with caution to prevent misuse and mistakes.

 




Popular Stories