Towards Human-Centric and Resilient Hospital Ecosystems: An AI-Driven Facemask Compliance Detection Framework for Healthcare Workers and Patients

Imagine walking into a bustling hospital. It's a place where science meets daily life, a high-stakes ecosystem where the goal is to heal, but where tiny, invisible germs can spread like wildfire if we aren't careful. To stop these germs, doctors and nurses wear special gear, like face masks, acting as a shield. But keeping everyone wearing their shields correctly is a huge challenge. This is where a branch of computer science called "Artificial Intelligence" (AI) steps in. Think of AI as a super-smart digital eye that never blinks. It uses a type of learning called "Deep Learning," which is like teaching a computer to recognize patterns by showing it millions of pictures, similar to how a child learns to spot a cat by seeing many different cats. The big question researchers are asking is: Can we build a digital guardian that watches over the hospital, instantly spotting anyone who forgot their mask, without needing a human to stand there and shout? This isn't just about following rules; it's about keeping the whole hospital ecosystem safe, resilient, and focused on the people inside it.

This paper is a massive "systematic review," which is a fancy way of saying the authors gathered and sorted through dozens of other studies to find the best digital tools for the job. They didn't build a new robot themselves; instead, they acted like expert detectives, comparing different types of AI "brains" to see which one is the best at spotting face masks in a hospital. They looked at several famous AI architectures—names like ResNet-50, MobileNetV2, and the YOLO family—treating them like different models of cars to see which one drives best on the bumpy roads of a real hospital.

The authors found that there isn't one single "magic bullet" AI that is perfect at everything. It's a bit like trying to find a single tool that is the best hammer, the best screwdriver, and the best saw all at once; usually, you need different tools for different jobs. However, they did discover a clear winner for the most critical areas. They found that models based on YOLOv5 (which stands for "You Only Look Once") are the heavy hitters for high-risk zones like Intensive Care Units (ICUs) and operating theaters. These models are incredibly fast and accurate, achieving precision and recall scores above 97% even when the view is blocked or the lighting is tricky. They are so good at handling "occlusion"—which is just a fancy word for when a mask is partially covered by a surgical loop or a hand—that they are the top choice for places where a mistake could be dangerous.

On the other hand, if the hospital needs something that runs on a small, cheap device (like a Raspberry Pi) or needs to work alongside temperature sensors, the MobileNetV2 architecture is the star. It's lightweight and fast, making it the most practical choice for general areas or for connecting to the "Internet of Things" (IoT), which is just a network of smart devices talking to each other. The paper also looked at other contenders like ResNet-50 and VGG-19. While ResNet-50 is a powerhouse of accuracy, it's a bit heavier and might be better suited for server-based monitoring in secure areas rather than real-time, on-the-go checking.

The researchers were careful to point out that simply having the AI isn't enough. They evaluated these systems against three big goals: Digitalization (can it work with the cameras we already have?), Resilience (can it handle the chaos of a real hospital, like bad lighting or people moving fast?), and Human-Centricity (does it treat patients, staff, and visitors fairly?). Their conclusion is that embedding these AI systems into the hospital's existing CCTV cameras can turn a manual, stressful job of checking masks into an automated, invisible safety net. This doesn't just stop infections; it helps the hospital bounce back stronger (resilience) and keeps the focus on human well-being.

So, what's the verdict? The paper suggests that by using the right AI tool for the right room—using the super-fast YOLOv5 for the critical care zones and the nimble MobileNetV2 for general areas—hospitals can create a "triple transformation." They can digitize safety checks, build a stronger defense against infections, and care for their people more effectively. It's a step toward a future where the hospital itself is smart enough to look out for us, ensuring that the only thing spreading is healing, not germs.