An ultrasonic body area network for implants

June 4, 2013

The body is roughly 60 percent water, which may make ultrasonic sensors a more efficient way to share information than radio-frequency signals (credit: University of Buffalo)

Researchers at the University at Buffalo are developing a “body area network” using ultrasonic waves and sensors to wirelessly share information between medical devices implanted in (or worn by) people to treat diseases such as diabetes and heart failure.

“This is a biomedical advancement that could revolutionize the way we care for people suffering from the major diseases of our time,” said Tommaso Melodia, PhD, UB associate professor of electrical engineering.

The idea of creating a network of wireless body sensors, also called a “body area network,” currently links sensors together via electromagnetic radio-frequency waves — similar to those used in cellular phones.

Radio waves have drawbacks such as the heat they generate, and because they propagate poorly through skin, muscle and other body tissue, they require relatively large amounts of energy, he said.

Ultrasound may be a more efficient way to share information, Melodia said, because roughly 65 percent of the body is composed of water. This suggests that medical devices, such as a pacemaker and an instrument that measures blood oxygen levels, could communicate more effectively via ultrasound compared to radio waves.

“Think of how the Navy uses sonar to communicate between submarines and detect enemy ships,” Melodia said. “It’s the same principle, only applied to ultrasonic sensors that are small enough to work together inside the human body and more effectively help treat diseases.”

Another example involves connecting blood glucose sensors with implantable insulin pumps. The sensors would monitor the blood and regulate, through the pumps, the dosage of insulin as needed in real time.

The project is investigating ultrasonic channel modeling and capacity analysis. physical/medium access control layer solutions for ultrasonic communications, distributed and asynchronous cross-layer control and resource allocation algorithms based on stochastic modeling of ultrasonic interference, and performance evaluation through a multi-scale simulator and a software-defined testbed.

The “Towards Ultrasonic Networking for Implantable Biomedical Device” project is supported by a five-year, $449,000 NatHional Science Foundation (NSF) CAREER grant. The CAREER award is the foundation’s most prestigious for young investigators.