Advanced devices developed by a mechanical engineering team at the University of Hong Kong (HKU) has proven to be useful for detecting potential stroke patients and helping machines mimic human brain functions.
In a collaboration with Nanjing University, Dr Paddy K.L. Chan, Associate Professor at the Department of Mechanical Engineering, developed a novel wearable electrocardiogram (ECG) sensor by integrating flexible, ultra-thin organic semiconductors into a flexible polyimide substrate.
Powered by a button battery, the sensor has outstanding signal amplification properties with a gain larger than 10,000, which allows it to detect electrophysiological signal, or f-wave with a frequency of 357 beats per minute (BPM), which indicates atrial fibrillation.
Conventional portable ECG sensors cannot easily detect the f-wave due to its weak amplitude. Atrial fibrillation is the most common arrhythmia associated with the increased risk of stroke or heart failure. The high signal detection capability stems from the ultralow subthreshold swing (SS) in the organic field-effect transistors (OFETs).
The study showed the ECG sensor managed to pick up unusual signals from patients with atrial fibrillation, while conventional electrodes could not.
He noted that people wearing the new sensors can also enjoy the freedom of movement, run around or even take a shower if they want, not being attached to a machine. A breakthrough in application with the use of a new device structure has been seen. The finding has been published in Nature Communications, in the article entitled “Sub-thermionic, ultra-high-gain organic transistors and circuits.”
Dr Chan’s previous breakthrough in developing the staggered structure monolayer OFETs, the material used in the latest experiment, was published in Advanced Materials. A US patent was also filed for the innovation. In the latest work, his team has advanced the application of the monolayer OFETs to flexible substrates for wearable electronic applications.
He said the subthreshold swing is an important parameter in transistor or inverter operation as it implies how much voltage change is needed to turn the device from an “off” state to an “on” state. The team’s devices provide a record low subthreshold swing device which ensures low operating power and high sensitivity.
The team also succeeded in adding ‘memory’ or collected signal, information to an organic transistor, which paves the way for advanced machine learning to mimic human brain functions.
The work has been published in Nature Communications, in another article entitled “Mimicking associative learning using an ion-trapping non-volatile synaptic organic electrochemical transistor”.
The team’s paper explains the physics behind how information can be stored in a device. It sets the stage for the next generation of computer learning through the enhancement of the ‘learning function’ of a device. For example, the memory transistors can be integrated with optical sensors for image processing and computation at the same time. The memory transistors are building blocks for the artificial neural network that can perform signal recognition or learn like a human brain, he said.
The team successfully added the “ion retainer” polytetrahydrofuran (PTHF) into a conductive organic polymer PEDOT:TOS. The PTHF can significantly slow down the move in-and-out of the ions in the PEDOT:TOS channel layer and maintain them at the desired conductance state. Multi-conductance levels, which can be considered as “memory levels”, were achieved. The experiment was held jointly with Northwestern University.
There is vast room for research in this area of human-machine interface, with unthinkable benefits for mankind. “There are unlimited possibilities when it comes to the applications of such interface,” added Dr Chan. In the meantime, however, he said that his focus would be on developing sophisticated circuits using advanced materials.