Facial recognition in the healthcare industryMarie Donlon | November 16, 2020
Facial recognition technology, the controversial technology that can match images of humans with a database of images to determine a person’s identity, is common in a host of industries from the electronics industry, wherein the technology is used to secure devices or as another layer of security in the transportation sector and virtually every industry in between.
Yet one industry where facial recognition technology is making inroads is an unexpected one: the healthcare industry.
Not only intended to thwart would-be criminals, facial recognition technology can play a budding role in the care of both patients and practitioners along with other healthcare-related applications. Following are a few examples of how the technology is being employed in the healthcare industry as well as future potential use cases for the technology.
Perhaps the most obvious role of facial recognition in the healthcare industry is to ensure the accurate identification of patients seeking medical attention. This form of digital identification could prevent the incidence of imposters seeking care under another person’s name and insurance. The technology could also be used to thwart security breaches, cybersecurity threats and fraud, improving a healthcare facility’s overall security.
Likewise, facial recognition technology could prevent the misidentification of patients, wherein the wrong patient receives treatment meant for another patient entirely. Experts suggest that the technology could reduce such treatment errors, and, consequently, legal action resulting from such errors.
Enable caretaking robots
As more and more segments of caretaking become automated in response to healthcare worker shortages, robots are gradually making inroads in the healthcare field. Naturally, robots capable of caring for patients are emerging, particularly during the global COVID-19 pandemic where both healthcare staff and resources are stretched thin and social distance mandates are being enacted. One way to better serve patients in hospitals is to outfit the caretaking robots with, among other advanced technologies, facial recognition.
An example of such a robot comes from researchers at Imperial College London who developed Robot De Niro, a robotics research platform that supports caregivers and interacts with patients.
Outfitted with facial recognition along with navigation and object manipulation, the Robot De Niro is capable of safely assisting and interacting with elderly populations.
Burnout among healthcare professionals is common, and more so in the midst of a pandemic where resources and clinicians are in short supply. Experts suggest that facial recognition might be used to eventually identify healthcare workers on the verge of burnout based on a combination of facial features and emotion detection. Identifying those employees and offering interventions early on could improve the employee’s state of mind and, consequently, the quality of care he or she gives to patients.
Prevent opioid addiction, drug-seeking
Paired with a pain medication pump or other similar dispensing equipment, facial recognition technology could potentially help administer pain medications to a patient following surgery or some other invasive procedure. The facial recognition technology synchronized with the pain medications could ensure that appropriate amounts of pain medication are administered based on the facial gestures of the patient, revealing when the patient is in most need of the medication. This could potentially be used to limit the amount of medication dispensed, reserving those amounts for only severe pain, and not issuing medication when pain levels are low as indicated by facial expressions. Thoughts are that administering appropriate amounts of medication would reduce the potential of subsequent opioid addition.
Similar to its use in identifying criminals, facial recognition technology could potentially serve to prevent the incidence of drug seeking that occurs in the healthcare industry, wherein those addicted to pain medication might visit multiple caregivers in search of replenishing hos or her supplies. Matched to a database of known drug seekers, facial recognition technology could easily thwart the efforts of a drug-seeking patient as soon as they set foot inside a medical facility.
Facial recognition is already being used to monitor patients in healthcare facilities. One example of how the technology is already in use hails from a team of scientists who are using facial recognition technology to predict when patients in a hospital intensive care unit (ICU) are engaging in potentially dangerous behaviors such as inadvertently removing their breathing tubes.
The tool, which was tested on postoperative patients in the Yokohama City University ICU, was devised using images captured by a camera mounted above patients’ beds. Roughly 300 hours of data were analyzed to locate daytime images of patients demonstrating good body positioning as well as clear views of those patients’ faces and eyes. A machine learning algorithm was then trained on nearly 100 of those images, enabling the model to identify high-risk behaviors, such as high-risk arm movements that could threaten the stability of breathing tubes, for instance.
Diagnose dangerous conditions or diseases
Facial recognition is already being used to help diagnose dangerous conditions and disorders. One instance of this is being conducted by a University of Colorado-led team of researchers. Using 3D-facial images, the team has reduced the amount of time it takes to diagnose a child with a rare genetic disorder based on how his or her features match up with those in a dedicated database. According to its developers, the technology reduces both the time it takes to reach a diagnosis and the number of medical tests the patient must undergo in order to be diagnosed.
Keep people healthy
As the COVID-19 pandemic continues to wreak havoc across the globe, many governments are requiring that masks be worn in public settings to prevent the spread of the virus. However, many people eschew the masks or wear the masks improperly, rendering them useless. As such, a team from the Chinese search engine company Baidu built a face-scanning tool to spot people in China, where the virus is thought to have originated, not wearing mandated face masks in public. The facial recognition technology was trained on about 100,000 images, enabling the face-scanning AI to eventually identify those not wearing the masks or wearing the masks incorrectly with an accuracy rate of 96.5%.
Despite being used in a number of different industries, facial recognition still has a long way to go before it can be used across all sectors without fear of misidentifications or inaccuracies affecting results. Until the AI algorithms used to develop facial recognition technology are built with “good” and diverse data, the technology will continue to be imperfect and potentially biased.
Check back with Engineering360 for more on the role of facial recognition technology in healthcare.