In hospitals and healthcare institutions, the sheer amount of patient metrics to track for the staff of doctors and nurses can be been a point of contention. Lawsuits based on the grounds of negligence are a risk that all healthcare practitioners take. Furthermore, there is an estimated 200,000 patients that die in the United States annually from medical errors. 1 Introducing wireless patient monitoring in these environments can potentially mitigate the risks that are innately involved in an environment geared at treatment and maintenance of sick people. There is always the potential for cyberattacks, but the rewards may outweigh the risks. Wired technologies limit patient mobility, increase the difficulty in transporting patients, and often introduce significant delays and hassle for the caregiver in arranging the cables.
With the patient monitoring market expected to hit $27 billion by 2020, wireless patient monitoring is poised to be a major part of that. According to Berg Insight, more than 3 million people worldwide were being monitored remotely by professional caregivers in 2013 — this is most likely from wireless medical services such as Medical Implant Communications Service (MICS) and Wireless Medical Telemetry Services (WMTS). WMTS is most often used for critical patient care whereas the more recently developed Medical Body Area Network (MBAN) technology can be used to simplify tracking biometrics in more routine patient care. Moreover, MICS and WMTS small bandwidths do not support high data rate applications.
Medical Body Area Network (MBAN)
In 2014, the FCC finalized the rules for MBANs — a network of sensors/actuators worn on the human body that communicate with a controlling device via a wireless link. With a spectrum allocation in the S-band from 2360 to 2400 MHz, the ruling states that the 2360–2390 MHz band is restricted to indoor use while the rest of band is open for use in other locations (e.g., residential). The MBAN is a subset of the more general trend of wireless body area networks (WBAN) or body sensor networks (BSN) that includes nonmedical applications such as human-computer interfaces (e.g., neural interface, virtual reality), location tracking, and personal fitness tracking).
In 2012, the IEEE 802.15.6 standard was released to define the physical (PHY) and medium access (MAC) layers of the WBAN Open Systems Interconnection (OSI) model. The standard defines wireless communications with support for quality of service (QoS), low power, and data rates as high at 10 Mbps while being able to operate in licensed (i.e., WMTS, MICS, MBAN), and unlicensed (ISM) spectrum space. 2 The PHY layer comes with three options: narrowband (NB), ultra-wideband (UWB), and human body communications (HBC) (see Figure 1).
Wireless Sensor Networks Paving a Path for MBAN Technology
Wireless sensor networks (WSN) are a nascent technology in the medical realm. Currently, WSNs are being heavily used in commercial applications such as smart homes with commercial off-the-shelf (COTS) transmit and receive modules such as ZigBee, Z-Wave, and XBee in the unlicensed 900 MHz and 2.4 GHz ISM bands. They are found in low power wide area networks (LPWAN) in both the unlicensed and licensed spectrum for long distance, smart city applications (e.g., smart lighting, smart meters, etc.). Industrial facilities have also used WSNs in industrial Internet of Things (IIoT) applications for tracking critical parameters around plants and manufacturing facilities.
The WSN attached to the body uses similar principles in a different piece of spectrum where sensor nodes are placed on and in various parts of the body for readings. In-body, or implantable, applications include glucose sensors, cardiac pacemakers, implantable cardioverter defibrillators (ICD), and endoscope capsules. On-body, or wearable, sensing applications include electrocardiogram (ECG), electroencephalography (EEG), body temperature tracking, blood SpO2 sensors, blood pressure sensors, respiratory rate monitors, spirometers, heart rate monitors, fetal heart rate monitors, and fall detection. By using relatively inexpensive sensor nodes, physiological information can be transmitted to medical servers in real time.
Factors to Consider for an MBAN
A key difference between an MBAN and other WSNs is mobility. While most applications are stationary, MBANs require the ability for the patient to move freely. This means the integrity of the wireless link must be maintained within a certain radius despite interference or even in non-line-of-sight (NLOS) conditions. Conveniently, the licensed spectrum space limits the risk of interference that comes with operating in the already congested ISM bands. This network will also most often be leveraged over short distances in hospital rooms, limiting the risk of interference.
Another factor to consider with any wireless device is battery life. The energy consumption is dependent upon the frequency of measurements taken and the frequency of transmissions (and received signals) from the sensor nodes. Compared to sensing data and low-level data processing, data transmission consumes most of the energy. 3 If the node is preprogrammed to notify the gateway in the case of a critical event, the battery may have a longer life as opposed to real-time data collection and transfer. For instance, a patient suffering from coronary heart disease may have regular ECG measurements taken at pre-allocated time slots throughout the day. The healthcare practitioner can then be alerted in cases where irregular patterns are detected, yielding a relatively small amount of measurements and transmissions. A case in which the battery would be drained rapidly may be with real-time monitoring of women with high-risk pregnancies where measurements and transmission are high in frequency. This type of intensive monitoring may benefit from energy harvesting methods to limit the maintenance of charging and changing batteries (see Figure 2).
The PHY layer would vary depending on the demand of the application. The higher the demand of the applications, the more throughput is necessary. The more throughput, the bigger the bandwidth. Narrowband technology with around 30 MHz of bandwidth would therefore offers lower power consumptions, longer battery life, and longer range due to an increased receiver sensitivity with a trade-off of data rate. Ultra wideband technology with 499.2 MHz of bandwidth offers higher transmission powers and higher data rates with a trade-off of battery life and shorter link distances. For example, a UWB transceiver would generate the power levels necessary to overcome propagation loss due to electromagnetic energy absorption in human tissue; a phenomenon known as the specific absorption rate (SAR). The data rates vary based on the sensor as well — a blood pressure sensor may require a low 10 bps on average while an endoscope capsule transmitting images can demand thousands of times more throughput on the order of 2 Mbps. 4
According to the FCC, the maximum transmit power for the 2360–2390 MHz band is 1 mW over a 1 MHz bandwidth primarily for indoor communications while the 2390–2400 MHz band has a maximum transmit power of 20 mW over 5 MHz for longer range residential communications. This network will most often be utilized in over short distances in hospital rooms, limiting the risk of interference. Longer link distances can be accomplished for the MBAN with the utilization of existing wireless personal area networks (WPANs) as a gateway, or the mediating device connecting to the cloud where the data can be collected and tracked.
Passive Hardware for WBAN Transceivers
The cost, power levels, and radiation pattern, as well as material and size constraints, are all factors that not only affect the active components of a transceiver, but also the passive. Cable assemblies and antennas vary based upon the application and environment. Current MBAN prototypes often leverage ZigBee or Bluetooth Low Energy (BLE) IoT modules that operate in the 2.4 GHz band. Factors to consider for the choice of antenna include gain and size constraints. Gain is a measure of the total power radiated from an antenna and is often plotted three-dimensionally or in two-dimensional slices as a radiation plot. The RF connector and cable are most likely going to vary based upon the size and environmental constraints. While there are many RF connectors that operate at 2.4 GHz, not as many come in small dimensions. Typically, ZigBee and BLE modules operate with a standard chip, PCB, wire or whip antenna along with a pigtail, or interconnect with a U.FL connector to mate to the PCB (see Figure 3).
Under the Bluetooth 4.0 standard, BLE is able to provide up to 1 Mbps data rates with extremely low power consumption with power idle modes such as standby and sleep where current can be as low as 1 μA. Therefore, this allows devices to operate for several months or years at a time with a single coin cell battery. Zigbee modules have nominal data rates of 250 kbps with as much as eight times the power consumption for a much longer link distance. For indoor applications within a short range, BLE may be a suitable option with a chip or PCB antenna. However, remote monitoring from a residence of elderly or chronically ill patients may likely require omnidirectional antennas with relatively high gain characteristics such as a whip antenna in order to link with local access points that are hardwired to the cloud via the Ethernet for tracking critical data. Smart assisted living scenarios such as smart homes equipped with acoustic and motion sensors may allow for mobile patient tracking where on-body and in-body sensors may often require antennas that are small. The sensors can then transmit information via a portable personal digital assistant (PDA) or mobile device where the patient data is transferred wirelessly over a wireless local area network (WLAN).
Chip antennas range in gain from 0 to 3 dBi and are more ergonomic since they can be directly mounted onto the PCB with the RF module. Still, they often exhibit a directionality where the radiating element performs better in certain orientations than in others. This characteristic tends to be less prominent in PCB and whip antennas but may be of no consequence depending on the distance of the link. This is a consideration for critical patients where data latency and reliability while the patient is moving are a high priority. While they do increase the dimensions of a sensor node, PCB antennas can mate directly to a Zigbee or BLE board via a U.FL connector and tend to have higher gain (2–5 dBi).
Connectors heads such as RPSMA, SMA, and MMCX are commonly used connectors that function in the 2.4 GHz band while fitting smaller form factor requirements. In environments where a patient is mobile motion, vibrational strain can potentially cause push-on connectors such as U.FL and MMCX to detach. The SMA and RPSMA threaded connectors have a more robust connection regardless of movement. Moreover, ingress protection (IP) may be necessary to ensure electrical performance over the lifetime of the device. An IP rating such as IP67 or IP68 can protect the connector mate from water immersion — this prevents the connected circuitry and subsequently the device from the water hazards that come with daily use.
Conclusion
The emerging fields of telehealth and telerehabilitation require a cross-disciplinary collaboration where medical and engineering minds can produce highly sophisticated systems for highly efficient patient care. Wireless medical spectrum allocations with WMTS, MICS, and MBAN can prevent interference with sensor nodes but still contain many considerations in patients that are mobile or are in a residential environment. Prototypes for 2.4 GHz MBANs often utilize ZigBee and BLE modules where it is critical to not only assess the software but also every connecting component in order to use these COTS devices for medical applications.
This article was written by Mark Miller, Product Manager, L-Com Global Connectivity, an Infinite Electronics company, North Andover, MA. For more information, Click Here .