DESIGN
EW: MEDICAL
essential to realising the full potential of Edge computing in MIoT devices. While performance is key, so is security. Healthcare data is extremely sensitive and is frequently subjected to malicious targeting. One study from Medigate and CrowdStrike revealed that 82% of healthcare organisations using IoT devices had been targeted by some form of cyberattack in 2020. As the number of MIoT devices increases, so does the potential attack surface for cybercriminals. Increased security measures to prevent unauthorised access, such as hardware encryption, secure boot processes, and tamper-resistant design, are essential. Mouser Electronics stocks a wide range of microcontroller units (MCUs), sensors, and hardware security modules to enable engineers to design smart and secure MIoT solutions. Analog Devices MAX78002 artificial intelligence microcontrollers (Figure 2) are designed to enable complex but localised AI neural networks. The MAX78002 can execute smart AI algorithms at ultra-low power, helping to elevate smart MIoT devices on the Edge. In applications like portable medical diagnostics equipment, the dedicated convolutional neural network (CNN) accelerator can allow for complex analysis and processing of locally captured data. With the hardware-based CNN accelerator, AI inferences can be executed in battery-powered medical applications using only millijoules of energy. The types of data analysed can vary depending on the type of medical equipment, but the MAX78002 is designed to be versatile and allows for a wide range of sensors to be connected to the 60 general-purpose I/O (GPIO) pins. It can also process visual information from the integrated 12-bit parallel and MIPI CSI-2 camera interfaces, with the CNNs capable of processing VGA images at 30fps. An additional I²S controller for digital audio also opens up the potential for intelligent sound-based medical diagnostics.
The MAX78002 addresses healthcare security head-on by incorporating a secure boot process, an AES 128/192/256 hardware acceleration engine, and a true random number generator (TRNG). When correctly implemented these measures can help better secure MIoT devices by reducing potential vulnerabilities. Edge Impulse To help streamline Edge development, Mouser has also established a strategic partnership with Edge Impulse, a specialist in embedded ML. The Edge Impulse platform empowers developers to quickly build, train, and deploy ML models directly to Edge devices, optimising performance and reducing the time needed to bring intelligent healthcare solutions to market. This solution supports the advancement of smart and low-powered healthcare solutions, like intelligent glucose pumps, which enhance insight, enable localised decision-making, and improve patient welfare. Conclusion Intelligent MIoT devices can address several key technical and operational challenges for often strained healthcare facilities, ultimately benefiting patients and society. The integration of Edge AI and ML-backed devices can revolutionise these environments by reducing network congestion, enhancing decision-making, and, most importantly, providing critical real-time insights that improve patient outcomes. As the adoption of MIoT devices grows the potential for smarter, more interconnected healthcare systems presents a clear solution to the challenges faced by healthcare institutions today.
Visit Mouser Electronics at embedded world: 4A-102
32 ELECTRONICSPECIFIER.COM
Powered by FlippingBook