Facial Recognition Revolutionizes Health Care Facility Safety and Infection Control

Workplace violence in hospitals and other health care facilities is increasingly on the rise, and the safety of staff and patients is paramount. In 2020, a meta-analysis study was done that reported that 61.9% of 331,544 staff members in the 253 studies examined had reported exposure to any form of workplace violence. One thing that hospitals can do is to prevent known threats from entering the facility before they can enter, and at the same time, only allow authorized personnel in secured areas of the hospital.

To find out more about this technology, recently, Infection Control Today^® (ICT^®) spoke with Dean Nicolls, chief marketing officer for Oosto, a technology company, Tel Aviv, Israel about their facial recognition technology. He discusses why facial recognition is needed in a hospital, the features that makes the technology unique and the company's future plans. He also explains how this product helps infection preventionists specifically beyond safety and into infection control and prevention.

ICT^®: Why is facial recognition needed in a hospital?

Dean Nicolls, chief marketing officer for Oosto: There are a few very important use cases. One use case is more hospitals are going touchless, and they're trying to avoid the spread of germs. For example, what I'm talking specifically about is access control. When I'm speaking about access control, that means who's allowed into different parts of the hospital. For example, in the operating room, you only want to have doctors and the patient. What facial recognition can do is ensure that the only people who get access to those very sensitive areas are the people who are given privilege to do that. It can also mean, even in other parts of the hospital, like the server room, you might only want specific IT (information technology) people in that server room to have access to that [because] there is a lot of sensitive information there.

That's one very important use case the hospitals are using facial recognition for, but there's a second emerging use case. That is to protect the staff, the nurses, the doctors, from any known threats. We're hearing increasingly in the news about hospital staff who are being attacked, in the hospital. These are people who have either an axe to grind because of maybe their attitudes about COVID-19. Or [they] could be a violent a spouse of a doctor or nurse. And the same technology that you use for facial access control, can be used to identify those known bad actors before they even set foot in the hospital, and make sure that security is alerted when they do set foot. And that's kind of an emerging use case to protect the hospital staff.

ICT^®: Please tell ICT^® about Oosto's facial recognition technology.

DN: It's important for your readership to understand facial recognition because it means a lot of different things to different people. The way that our facial recognition technology works with the existing camera system inside the hospital. The cameras are streaming from their CCTV cameras—and they have cameras spread throughout the hospital and sometimes in the garage and the front entrances…operating rooms and other sensitive areas. We're essentially analyzing the camera stream in real time. We're looking for is a face in the frame of the video. And again, think about how fast this is happening. This is all done via artificial intelligence (AI) and neural networks. Once we identify that yes, there's a face, we then look for the distinguishing characteristics of the face. Then we're comparing that, essentially creating almost a facial map of that person's face in real time, and then comparing that to a known list of people who are on a watch list, or could be a list of authorized employees if people who are on that. For example, in the example I gave of a dangerous person, as people are walking through the front entrance, even if they're in a crowd, or they're in a group, we're analyzing each of the people in that group, say is any of these people on the watch list? And if they are, we're then notifying security automatically for them. So, [security] can then take appropriate action.

The way it works with access control, [is] just the reverse. You might have a list of 1000 employees who are meant to be in the hospital. And then in many of the entrances to the hospital that are for employees only, you might have the employee look at a little reader. looking at a camera. And just by looking at the camera, we know this is an authorized employee. They're on the good list, and then the doors will open. And again, they can do that without having to touch any surfaces or any keypads. So that's part of the touchless story.

And if you think about doctors who have just scrubbed and they go into the operating room, they can enter an operating room by just looking at a sensor and say, “Oh, this is the doctor,” even if he has a mask on. Then because we can recognize that person as being on the authorized list, the doors open, and then the doctor can enter. And anyone who’s not authorized and nonemployees, we don't recognize this given employee, so the doors don't open. So that's the way of keeping kind of the riffraff out. It's also a sense of ensuring that only the right people are where they ought to be.

ICT^®: What is the failure rate for this technology

DN: It is surprisingly low. It is less than a percent.

ICT^®: What are the specific features that makes this technology so unique?

If you think about access control, historically, the way the hospital would deal with access control is they would use probably one of three methods who use like a car key or a fob. You have to carry this car key with you. And you swipe it against a panel on a wall, and that gives you access, or you touch a keypad and you put in a code. For the last two decades, most hospitals use some kind of system like that. Or if they had a biometric system, they might use a thumbprint or something like that. But as we've gotten touchless right, there's been a need to move away from those technologies. The other problems with some of those technologies like the card keys and the fobs is that they can be lost, stolen, or shared. I could share my card with you and you might be a bad actor. You now have access to the hospital. You can access any part of the hospital and most hospitals are concerned about it. They only want the right people access in a hospital. They're trying to close those security loopholes. Using biometrics like the face is an emerging way that a lot of modern hospitals are now evaluating, not only because you can close those security loopholes, but because you can also allow your hospital to go touchless.

ICT^®: How does this product helped infection preventionist specifically?

I think the primary reason is the touchless area. The notion that you can go from room to room within a hospital and specifically going where you get scrubbed and clean into the operating room. And to use this technology without having to touch anything. Because you've just been scrubbed. Some of the examples we saw on that footage in thehospitals [on the Oosto website], or doctors holding their hands like this [hands up after preparing for surgery], as they go through the door, and doors are automatically opening; they don't need to touch the doors? The ability to even recognize that in certain areas, you might need all personnel to be wearing a mask. We can use this technology to alert people. If someone enters an area that's supposed to have everyone have a mask, and they don't have a mask, we can alert [the proper] people that the person isn't wearing a mask. These are the areas a lot of it in your specific area that we're referring to the access control going touchless is the primary benefit of this technology.

This interview has been edited for length and clarity. Watch the video for the entire interview for additional information.