Designers guide: Sensors for medical devices

The healthcare industry is progressively moving from a centralized, clinical model to a more patient-centric approach, requiring monitoring solutions that are portable, wearable, and patient-focused. This process involves significant technical and hardware challenges. Designers must find a way to maximize diagnostic accuracy and reliability in a clinical setting while also keeping power consumption, size, and long-term reliability in mind.

The design of a medical device usually follows a modular approach. This means that each part, from the first signal capture to the last communication protocol, must be optimized for speed and efficiency. This article will provide insights into some of the most relevant sensors employed in medical devices as well as the associated technologies, including analog front ends (AFEs), power management devices, and wireless system-on-chips (SoCs) for connectivity.

Pressure sensors

With the wide range of medical devices, from wearable glucose monitors to computerized tomography (CT) scan equipment, there is also a variety of sensors incorporated into these devices. These include pressure and temperature sensors as well as biosensors and accelerometers.

Pressure sensors are used in a wide range of medical equipment, from non-invasive blood pressure monitors to specialized airflow sensors in ventilators. Besides common medical requirements, such as reliability and high sensitivity, these sensors must provide robustness and endurance. Medical-grade pressure sensors must exhibit high linearity and long-term stability, maintaining calibration over weeks or months of continuous operation.

For example, TDK Corporation offers a wide portfolio of piezoresistive pressure sensor dies well-suited for high-precision measurements in the medical sector. Based on advanced silicon MEMS technology, these sensors are grouped into three main categories: absolute, gauge, and differential pressure.

As for the piezoresistive pressure measurement methods, sensor dies are available with frontside and backside absolute measurement and gauge-differential measurement. The frontside configuration, where the electronics are directly exposed, is preferred for dry, non-aggressive gases. The backside design allows the sensor to handle wet media or non-aggressive fluids because the sensitive electronic components are shielded on the opposite side of the pressure-sensing diaphragm. Finally, the gauge configuration is well-suited for physiological measurements relative to ambient.

The C39 series are highly miniaturized dies (with an area of 0.65 × 0.65 mm) with frontside absolute pressure measurement up to 1.2 bar. Able to operate over a temperature range from −40°C to 150°C, these sensors are optimized for high burst pressure and feature narrow sensitivity tolerances and high signal stability. As such, they are suited for integration into high-density wearable medical devices.

Sensors for medical imaging

Imaging technology has made significant advances in the last few years. CT scans used for the diagnosis and monitoring of various conditions, including cancer and cardiovascular diseases, have evolved with the introduction of the photon-counting CT (PCCT).

The main difference between these two techniques lies in how sensors (“detectors”) process X-rays. Conventional CT uses indirect energy-integrating detectors. X-rays hit a scintillator, convert it to light, and then to electricity. In practice, they measure the total energy accumulated, losing individual photon data.

PCCT instead uses direct-conversion sensors that convert X-rays directly into electrical pulses, counting every single photon and measuring its specific energy. This eliminates electronic noise, improves spatial resolution, and allows for precise tissue differentiation at a lower radiation dose.

Ams Osram, now part of Infineon Technologies AG, introduced a system-in-package (SiP) sensor module specifically designed for photon-counting detectors. This sensor, shown in Figure 1, enables a significant reduction in the radiation dose and diagnostic images with higher resolution.

As the company states, the AS5920M module features a 9× reduction of the module’s detector pixel size compared with traditional CT systems. Moreover, more modules can be combined in an array arrangement, increasing the detection area according to the desired CT application.

At the 2025 annual meeting of the American Society for Radiation Oncology, Siemens Healthineers presented the Naeotom Alpha.Prime PCCT scanner (Figure 2) based on cadmium telluride crystal detectors that significantly improve image resolution and contrast. The company introduced the world’s first PCCT scanner in 2021.

Embedding AI in sensors

The integration of embedded AI cores directly into biosensors is changing the architecture of medical diagnostics. Previously, devices were limited to a traditional sensing process, wherein all raw data was transmitted to a central processor for analysis. With the direct integration of edge intelligence, sensors can now process data locally, exactly where it is sourced.

The main benefit of this architecture is efficiency, as the device transmits only processed results or alerts. This approach significantly reduces the system’s power consumption, latency, and required bandwidth.

STMicroelectronics has introduced a high-accuracy biosensor that integrates a vertical AFE (vAFE) for biopotential signals (typically cardio and neurological parameters) with a low-power, three-axis accelerometer with AI and anti-aliasing. The ST1VAFE3BX’s vAFE features programmable gain and input impedance and includes a 12-bit ADC.

Providing output data at a rate up to 3,200 Hz, the biosensor is well-suited for biopotential measurement of heart, brain, and muscular activities. The compact size (2 × 2 mm) and reduced power consumption (48.1 µA during normal operation, which can be cut to just 2.6 µA in power-saving mode) suit it for wearables designed for predictive healthcare.

The biosensor features ST’s proprietary machine-learning core (MLC) and finite-state machine (FSM), which allow designers to develop decision-making rules and algorithms to be deployed directly on the chip. The AI-assisted capabilities enable the sensor to autonomously manage motion and activity detection.

This AI feature decreases the interactions with the host controller, reducing the overall power consumption and latency while extending battery life. MLC and FSM can be implemented using ST’s software development tools such as MEMS Studio, which is part of the ST Edge AI Suite.

High-precision AFE

The integrity of a medical device is defined by the quality of its input data. AFEs are components required for interfacing with the human body. They are essential for all types of medical sensors that produce analog signals and therefore require further processing, such as conditioning, amplification, filtering, and digital conversion.

AFEs bridge the gap between physical measurements, typically available in analog form, and the compute device that processes them in digital form. In medical devices, AFEs are required for any sensor that measures physical parameters.

AFEs operate by extracting small-amplitude physiological signals from the environment, which are often noisy or subject to electromagnetic interference. As a result, to achieve medical-grade results, the AFE must provide a high signal-to-noise ratio and low leakage currents.

Among the sensors that require an AFE are biosensors, such as those used in continuous glucose monitoring (CGM) and electrocardiogram patches. Onsemi’s CEM102 is an AFE specifically designed for CGM and similar applications. Based on an amperometric measurement that senses very low currents, the device features a small form factor and low power consumption. These features suit the CEM102 for miniaturized and battery-operated medical devices.

The CEM102 can be operated with a supply voltage ranging from 1.3 to 3.6 V—typically a single 1.5-V silver oxide battery or a standard 3-V coin cell. It supports up to four electrodes, integrates a high-resolution ADC and several DACs for bias setting and a factory-trimmed system, and can be interfaced with a host controller, such as the onsemi RSL15, a secure Bluetooth 5.2 wireless microcontroller (MCU) for connecting to an external device or terminal.

Power management

In the design of compact wearables, such as hearing aids, power management represents one of the most challenging constraints. Designers must select power management integrated circuits (PMICs) with high efficiency, thus preserving the energy provided by small battery cells.

Onsemi’s HPM10 battery-charge controller is a high-performance PMIC engineered to recharge batteries in miniaturized medical devices, typically hearing aids and cochlear implant devices. The device supports different rechargeable battery technologies, including lithium-ion and silver-zinc, and can detect zinc-air and nickel-metal hydride disposable batteries.

The HPM10 also provides a charger communication interface to communicate the state of the charging process to the hearing-aid charger. Other information available on this interface includes the battery voltage levels, current levels, temperature, and battery failures.

Connectivity

A medical device is more effective if it can communicate data to clinicians or electronic health-record systems. Several connectivity protocols are available, and their selection is based on the application’s range and data throughput requirements.

Low-power Bluetooth SoCs are the industry standard for wearables, providing a reliable and efficient link to smartphones or home gateways. For high-bandwidth clinical environments, such as hospitals or clinics, integrating Wi-Fi 6 with Bluetooth Low Energy (LE) represents a suitable connectivity solution.

For example, Silicon Labs’ Series 2 BG29 family of wireless SoCs is designed to provide Bluetooth LE connectivity in an extremely small form factor. The BG29 device’s small size (2.6 × 2.8 mm) suits it for applications such as wearable health and medical devices and battery-operated sensors. The device integrates a DC/DC boost converter supporting a wide voltage range, a Coulomb counter for accurate battery monitoring, 1 MB of flash, 256 kB of RAM, and security features.

NXP Semiconductors is collaborating with Silex Technology, a provider of wireless connectivity and smart edge solutions for the medical and industrial sectors. Silex focuses on wireless solutions for medical applications requiring high longevity, cybersecurity features, and high reliability. Patient monitors, medical wearables, and other connected devices often operate in hospitals where several Wi-Fi access points are available.

Silex integrates NXP’s Wi-Fi SoCs in its Wi-Fi 6 + Bluetooth 5.3 and 5.4 module solutions, including NXP’s IW611 Wi-Fi 6 SoC and RW610 Wi-Fi 6 wireless MCU.

The post Designers guide: Sensors for medical devices appeared first on EDN.