Get email delivery of the Cadence blog featured here
One powerful countermeasure against chronic disease is to detect and treat it early. Doing so calls for collecting and monitoring vast amounts of data. In Dr. Jerald Yoo’s view, this is the perfect job for wearable healthcare devices.
“Wearable electronics is needed for proactive healthcare,’’ Yoo said during a recent lunchtime talk arranged by the Cadence Academic Network at Cadence’s San Jose headquarters. Yoo is an associate professor in the Department of Electrical Engineering and Computer Science at the Masdar Institute of Science and Technology in Abu Dhabi. During his talk, Yoo covered the challenges and techniques to designing biomedical circuitry, using a closed-loop seizure detection microsystem as his example.
According to the World Health Organization, about 50 million people worldwide have epilepsy. Currently, diagnosing this severe neurological disorder involves doctors interviewing the patient and administering an electroencephalogram (EEG) test, said Yoo. But these methods are hardly conclusive—continuous monitoring is needed, he said.
With existing health monitors, such as those for blood pressure, arrhythmia, and epilepsy, patients feel encumbered by all of the wires. Easy-to-use, wearable health monitors that are developed on a flexible platform are an answer, says Yoo.
Design Considerations for Healthcare Wearables
What are the design considerations for wearable health devices?
On the platform side, the introduction around 2009/2010 of printed fabric circuit boards was quite the revelation. Direct screen-printing of conducive ink on fabric has brought to life many wearable applications. The technology also provides an alternative to wet electrodes (which can trigger skin sensitivities if worn for long periods) and dry electronics (which have high electrode impedance and, thus, more noise). Designing fabric circuit boards comes with its own challenges, from the pad number limits to the need to address heat protection, static and dynamic parameter variation, and high impedance.
For the sensor I/F circuit, it’s important to have a dedicated DC server loop to remove the electrode offset. Since the servo loop itself elevates noise, Yoo and his students created a design that includes a 500Hz chopper at the servo loop for better noise efficiency.
Considering the digital backend, there are some distinct EEG seizure detection challenges to consider. Namely, intra-patient age-to-age EEG variations and spatial EEG variations are unexpected outcomes. How to solve this? The introduction of machine learning via support vector machines (SVMs) provides an answer. There are two options here: linear SVM, (LSVM) which requires limited seizure patterns but offers moderate classification accuracy, and non-linear SVM (NLSVM), which requires sufficient seizure patterns and has high classification accuracy. Yoo’s choice was to use two LSVMs, one trained for sensitivity and the other trained for specificity. With this approach in a single system, he found accuracy rates of 95% for sensitivity and 98% for specificity detection performance.
Lastly, there are design considerations around the seizure detection system, which is essentially a wirelessly powered ECG. For this, Yoo and his students used a fully integrated EEG SoC consisting of an analog front-end and digital backend with SVM and simulation, scalable EEG processing, and machine learning for patient-specific seizure detection. All of their work was implemented using Cadence tools.
In summary, Yoo noted, to be effective for proactive healthcare, wearable electronics must be minimally obtrusive, disposable, and energy efficient.