Chinese scientists develop ECG radar that can detect heart health from a distance

Most ECG devices depend on skin sensors, but a University of Science and Technology of China team has developed a contactless alternative. Radio wave scanner can detect cardiac motion and transform the data into electrical signals with 90% morphology accuracy, study says. — SCMP

Chinese scientists have developed a radio wave scanner that can detect the heart’s electrical activity from a distance, making remote monitoring of vital signs possible for the first time.

Regular ECG monitoring is known to play a crucial role in diagnosing and preventing heart disease. However, most such devices require sensors, or electrodes, which are placed on the patient’s chest area to record electrical signals from the heart.

Although this allows for efficient monitoring of cardiac health, the need for skin attachment can “restrict the reliability, adaptability and continuity of monitoring”, researchers from the University of Science and Technology of China said.

Chinese scientists used a millimetre wave radar to detect the movements of the heart and transform that information into electrical signals. Photo: The University of Science and Technology of China

In their new study, the scientists explained how they used a millimetre wave radar to detect the movements of the heart and transform that information into electrical signals. This contactless ECG system could help in the daily monitoring of certain heart diseases, the team said.

It is the first electromagnetic wave-based device to exploit the relation between the heart’s beats and its electrical activity, according to the authors of the study – published in the peer-reviewed journal IEEE Transactions on Mobile Computing in October.

Past studies have demonstrated that radio frequency-based sensing can monitor the heart’s pumping cycles, but solely from its mechanical activity perspective.

The authors designed a series of signal-processing algorithms to extract the 4D cardiac motion signals in order to comprehensively describe the heart’s mechanical activity.

They first separated the readings coming from other parts of the body, eliminated motion interference such as from breathing, fine-tuned the signals from cardiac movement and suppressed ambient noise, before finally extracting the data.

The authors then devised a data-driven neural network to transform the radio frequency input into electrical signal output.

The team collected 10 hours of contactless and traditional ECG data from 35 study participants to compare the accuracy of the readings.

The study found that the contactless radar was less than 14 milliseconds behind in time accuracy, while it showed more than 90% of morphology accuracy.

Corresponding author Chen Yan said the results supported the stable monitoring of the interbeat interval – a key indicator in the diagnosis of heart disease – within 9 milliseconds of error.

“We are currently working with relevant hospitals and once the technology obtains clinical acceptance, it will provide an important aid in the daily monitoring and diagnosis of arrhythmia, heart attack and other diseases,” Chen was quoted as saying by the official China Science Daily. – South China Morning Post

Article type: free
User access status:
Subscribe now to our Premium Plan for an ad-free and unlimited reading experience!

Next In Tech News

Uber to steal a march on Lyft in resurgent rideshare market
Meta settles U.S. lawsuit over infinity-logo trademark
U.S. judge rejects new bail conditions for FTX founder Bankman-Fried
Consumers to face 20,000 pound limit on digital pound, Bank of England says
U.S. prosecutors ask to postpone SEC, CFTC cases against Bankman-Fried
Ex-Coinbase manager pleads guilty in insider trading case
Central bank test lab to trial 'stablecoin' monitoring system
Exclusive-AFL-CIO wants bigger say in U.S. digital trade deals for workers
Retail investors flock to small-cap AI firms as Big Tech battles for share
STMicro leans on AI, cloud as chip designs become more complex

Others Also Read