The following is a guest blog post by Yaroslav Kuflinski.
As AI, or artificial intelligence, gets more traction across the healthcare realm, it practically becomes the center of innovation in the industry. Vendors and clinical stakeholders base their solutions around AI to optimize numerous care and administrative processes, including workflow assistance, diagnosis facilitation, and clinical trial support. One of the recent and most promising additions to the AI family is facial recognition software.
Facial recognition is the next level of digital identification, a step further from more fragmented approaches to identity verification, such as recognizing unique patterns of palm veins, scanning iris, and reading fingerprints. According to a recent report from MarketWatch, the global facial recognition systems market is anticipated to hit over $1.1B by 2024, growing at a 13% CAGR between 2019 and 2024.
Given that healthcare is well-known for being under constant pressure from both physical security breaches and cyber threats, the industry primarily invests into facial recognition to enable more advanced facility security and fraud prevention. Yet, this technological approach can also be used to improve care delivery and even support physicians’ mental health.
Automating Fraud Prevention
If face recognition technology is applied to a video surveillance system across medical facilities, it can become a non-intrusive yet effective way to detect and identify suspicious people like those previously flagged or wanted by authorities, drug seekers, insurance imposters, and others.
Such a system will be able to identify patients as well as medical specialists, while new visitors can be easily sorted into a separate category. In case a person is checked and identified as a potentially fraudulent individual, the facility’s security team will be able to mitigate the situation without drawing too much attention.
For example, if a drug seeker or a person who tries to pose as a patient is in the facility, the security specialists can notify a few administrators and make them isolate the person, calling them into ‘a doctor’s office’. From there, a person can be hеld up for further investigation according to internal policies.
Streamlining Patient Identification in Acute Cases
While cybersecurity threats can entail costly consequences if patient data is leaked, most don’t pose direct risks to patients’ lives and their post-treatment health outcomes. Medical errors do, however.
If the patient is wrongly identified in a healthcare IT system for some reason, this may lead to misdiagnosis, inadequate site procedures, incorrect therapy administration, and other unintentional but potentially harmful actions. Facial recognition software can be used as an additional safeguard measure and an interim step in the patient identification process.
Facial recognition significantly facilitates patient identity verification, and while it can be a valuable addition to other biometric analysis methods, it can excel where others fail. For instance, a person can lose their RFID bracelet or another location tracker, they might be unconscious at admission, or using a fingerprint scanner can be undesirable due to infection transmission risk in particular cases.
At the current AI maturity level, algorithms are able to achieve more efficiency in recognizing patients live than when verifying them by images, even those made in close-to-perfect conditions.
Now AI can also account for variations in a person’s look, such as different angles, lighting, hairstyles, and facial hair. Accordingly, the algorithms are able to succeed in identity recognition even when a patient can’t name themselves or focus their sight on the camera, like when a person is admitted after a physical injury, in distress, or in an unconscious state.
In a recent publication, a team of researchers tested out their proprietary facial recognition algorithm for the pediatric care setting. They specifically tried to carry out patient identity validation in cases where the study participants weren’t able to look into the camera or their facial features were slightly distorted, for example, children in an unconscious state, under anesthesia, or after surgery. The researchers achieved a 99% accuracy in identifying both inpatients and outpatients.
Detecting Emotions in Real Time
Apart from identifying people, face recognition technology also enables real-time emotion detection. This feature can bring in multiple benefits to healthcare organizations, both in care delivery and internal operations.
For instance, facilities can fight emotional burnout among health specialists. According to a recent survey by Medscape, 44% of physicians reported feeling burned out. Moreover, 4% of them were diagnosed with clinical depression and 11% were colloquially depressed, meaning that they were experiencing sadness, anxiety, and indifference.
By applying facial recognition to emotional state analysis, medical organizations will be able to independently assess specialists’ frustration levels and build a system for burnout prevention.
For example, providers could optimize working and vacation schedules for the staff members with the most disturbing readings, introduce meditation practices, or reward employees with any other recreational activities. This way, organizations can make sure that every care team member is at the top of their productivity and in good mental health, so they can focus on patient needs better. Such practice can also lead to an overall decrease in medical errors caused by human factor.
Optimizing pain relief
Real-time emotion detection can also be applied in care delivery, namely, in pain management. Sometimes, patients recovering after a serious injury or surgery may not be able to use the patient-controlled analgesia pump to control the flow of painkillers. Additionally, some patients can misuse the pump, risking to develop an opioid addiction.
With real-time emotion analysis synchronized with the pain medication pump, the system can capture the slightest changes in patients’ facial expressions, calculate and deliver the exact doses needed to ensure patients’ comfort without the risk to trigger an addiction.
Reading Faces, Saving Lives
Many industries, including retail, banking, entertainment, and sports, invest in facial recognition to get a competitive advantage in transaction processing and service improvement.
But when it comes to healthcare, the reasons to adopt facial recognition have more life-changing implications. When clinical stakeholders hop on the next big thing, they put the business potential last and focus on the practical benefits for their staff and patients first.
Summing up, facial recognition can become just the right technology to make the difference across the healthcare realm. However, only time will tell.