By Warren Miller, Mouser Electronics
Many portable medical devices are already familiar to us. Digital thermometers, blood pressure monitors, blood sugar monitors, blood oxygen meters and pulse/heart rate monitors are all familiar non-invasive medical devices. Over time these devices have migrated from the doctor’s office to the home. This trend continues as devices are being created with ever more sensitive monitoring capabilities and advanced algorithms to allow diagnostic and treatment capabilities to migrate outside of the doctor’s office.
A key element in this migration is the addition of intelligent guided operation instructions that make it easy to use instruments that are considered very complex. For example, even today some medical devices can help in emergency situations, even when a medical professional isn’t available. These devices can ‘stand-in’ for a doctor or nurse and provide a degree of care not previously possible outside of a hospital setting. By looking at the design of one of these intelligent medical devices we can better see the direction this new wave of intelligent medical devices will evolve and the types of capabilities we can expect.
Intelligent Operation – The Automated External Defibrillator
An Automated External Defibrillator (AED) is one of the most obvious intelligent portable medical devices, and one often seen in airports, conference centers and public buildings. An AED is used in an emergency to respond to sudden cardiac arrest (SCA). It can use an electric shock to restore the heart back to normal operation. Furthermore, AEDs found outside of a medical setting are designed to be operated by a ‘lay’ person, not a medical professional. These devices can provide simple audio and visual instructions and guidance to the operator that walks them thru the procedure for set-up, diagnosis, treatment and even alerting emergency medical services. The device automatically takes readings, makes diagnoses, and determines the most appropriate action to take.
Figure 1: Example AED Device. Image courtesy of Defibtech
The Design of an AED
Perhaps the most illustrative aspects of the design of an AED that also apply to other medical devices as well are the analog sensing and filtering, digital signal processing, human interface and communications functions. By looking at the design of these key functions in some detail, techniques used in the next generation of portable medical devices should become clearer. In order to better understand the sensing, diagnostics and operations, the AED will need to perform a quick summary of how the electrical system that ‘powers’ the heart works.
Monitoring the Heart
The two most typical cardiac arrhythmias that an AED device can currently address are ventricular fibrillation and ventricular tachycardia. In both of these AED-treatable conditions, the heart is active, but operates in a dysfunctional pattern and may progress to cardiac arrest. In ventricular tachycardia (V-Tach), the heart beats too fast to pump blood effectively. Prolonged ventricular tachycardia, if not treated, leads to ventricular fibrillation (V-Fib). In ventricular fibrillation the heart controlling electrical activity becomes chaotic, interrupting the ventricles normal efficient pumping of blood. This fibrillation in the heart decreases over time, and will eventually reach a failing ‘flat-line’ or asystolic state. In order to better understand the design of a heart rate monitor, a quick review of the electrical operation of the heart is useful.
As the electrical activity is spread throughout the atria, it travels from the SA node to the Atrioventricular (AV) node. The AV node functions as an important delay in the conduction system and is shown as the PR segment of the ECG plot. The QRS portion of the ECG shows the rapid depolarization of the right and left ventricles. Because of the much larger muscle mass in the aorta, the QRS complex has a much larger amplitude than the P-wave. The T-wave shows the re-polarization of the ventricles. The duration of the P-wave is on the order of 80ms, the QRS complex from 80ms to 120ms and the T-wave lasts about 160ms.
Figure 2: Heart Structure and Electrocardiogram of Normal Heart Operation.
For Heart Structure see page for author [GFDL or CC-BY-SA-3.0], via Wikimedia Commons
Electrocardiogram Created by Agateller (Anthony Atkielski), converted to svg by atom. [Public domain], via Wikimedia Commons
In most AED devices two sensor pads are used to measure the heart’s electrical activity. These sensors can also function as the source of the electrical shock used to change an irregular heartbeat back into a regular, safe rhythm. During sensing, the pads measure the voltage generated by the heart during the polarization and depolarization of the components of the heart. These voltages are very small and are typically on the order of 10mV. Because the signal will vary from individual to individual and will depend on the placement of the sensors, accurate measurement is critical for the most effective device operation. A high performance Analog-to-Digital Converter (ADC) is required to make sure the signal is captured as accurately as possible. The frequency of capture is not very high, since we are dealing with a physiological process, so an ADC that operates audio frequencies should be more than sufficient (remember the time periods of the various waves in the ECG were around 80-160mS). Modern MCUs typically have on-chip ADC peripherals with sufficient capabilities to process audio frequencies.
Some medical equipment may require more complex measurements so the use of an analog front end (AFE) might be required. In wearable devices, like fitness and medical monitors, a new technique that uses LEDs for pulse oximeter and heart rate sensing has been developed. This technique will likely find its way into other portable applications as well due to its low power and small size. The Maxim MAX30100, for example, integrates the sensor and LED in a single device. The sensor features a high signal-to-noise ratio to remove motion artifacts, high sampling rate, and fast data output, making it a convenient AFE for an MCU.
Figure 3: Maxim MAX30100 Pulse Oximeter and Heart-Rate Sensor with Integrated LED and Sensor.
AEDs require multiple filters for signal processing, as do most intelligent medical devices. The first filtering operation is used to eliminate background noise from the critical signal components. A high-pass filter is often used to remove irrelevant low frequency signals and might typically be set at 0.05Hz in an AED. A low-pass filter is used remove irrelevant high-frequency signals and for an AED might be typically set at 150Hz. Other medical devices might use different filter settings, but just about every device would need to set filters to ‘window-in’ on the portion of the signal that is most important for the diagnostic algorithm.
Once the signal has been isolated from any noise sources, key measurements can be made of the critical timing and amplitude characteristics of the captured ECG waveform. The measurement of the time periods and amplitudes for the P, QRS and T sections of the ECG are important to establish the rhythm (or lack if it) as the cardiac event begins and develops. Other advanced diagnostic measurements can be determined to further improve operation. For example, the direction of the heart’s mean electrical vector (as opposed to just the amplitude) can be used to determine if a blockage or if heart disease exists; important factors in a correct diagnosis. A properly operating AED will not discharge if it does not sense a heart rhythm that requires it. Ostensibly, an AED is to be used by trained personnel. However, if the alternative is possibly death, it’s worth it to follow the AED instructions, especially if directed to do so by (possibly remote) medical personnel.
Because of the chaos and uncertainty involved with a cardiac arrest it is important for the AED to take the operator through the use of the AED in clear, simple and easily implemented instructions. One of the most helpful features to assist a user is a series of audio prompts, spoken in a confident and reassuring voice. The prompts should be in the operator’s language, perhaps selected from a touch screen prior to the start of operation. Video instructions can be used to augment audio instructions and the availability of high resolution LCD panels that interface easily with MCUs makes this an easy feature to include. The large amount of storage available in serial flash memory devices can easily provide a variety of video instructions covering a wide range of possible scenarios for diagnosis and treatment procedures. Serial flash can also store audio and video captured during treatment to help first responders, to check that audio and video prompts were followed and for review by hospital staff to better determine the proper follow-up care.
Communications Capabilities Further Improves Emergency Response
Intelligent monitoring and communications between an operator and emergency services can also improve treatment outcomes. While stored in the easily visible kiosk, the AED should be charged and periodically function tested, and if any test fails the AED could alert a service organization to replace the device. If someone tries to remove a device which has failed the self test, the AED could use a local communication network to automatically find the closest working device and automatically direct the operator to that unit. This can be of critical importance during a significant emergency event such as an earthquake or fire where devices might be damaged or lose power.
The next wave of portable medical devices deliver intelligent diagnostic and treatment options previously only found in the hospital or doctor’s office. These devices will be easy to use at home, or in the case of emergencies, in public buildings and offices even by untrained professionals. It won’t be long until the answer to the question, “Is there a doctor in the house?”, will always be “Yes”. And the doctor will be you.