< Back to previous page

Project

Ultrasound Waves Based Body Area Networks

Implanted or ingested medical devices aid in the localized diagnosis or treatment of several major diseases. For example, the cochlear implant stimulates an electrode in your cochlea, or the endoscopic pill visualizes the internals of your digestive tract. These medical devices typically include a wireless in-body communication link to enable remote monitoring, configuration, and control. This communication link needs careful optimization to facilitate a reliable and secure data link, combining high communication throughput with low energy consumption.

Different in-body communication technologies exist, stemming from the need for diverse data rates and implantation depths. In-body ultrasound (US) wave based communication benefits from a lower human body attenuation compared to electromagnetic waves and intrinsically provides a more secure communication link because the ultrasound waves stay within the body. Moreover, it features a safe technology highlighted by the preferred way of imaging a child in the mother's womb. Compared to other in-body communication technologies, the ultrasound wave's lower channel attenuation and shorter wavelengths promise an enhanced design in the trade-off triangle formed by the transducer/antenna size, the communication data rate, and the in-body communication distance.

However, state-of-the-art ultrasound-based systems employ large, bulky transducers not suitable for implantation, or they lack a hardware proof of concept in an energy-constrained setup. Furthermore, they often apply a non-multipath-tolerant communication modem, whereas the human body contains many ultrasonic reflective materials such as bone and air bubbles.

This work aims to advance the in-body communication link distance, reliability, and miniaturization through ultrasound wave technology. More specifically, it focuses on symmetric transceiver applications, such as those found in wireless in-body networks where an implanted sensor communicates with another implanted processing unit or stimulator device. In such applications, the data rate is moderate (0.1-1Mbps) while requiring an in-body communication distance beyond 10cm.

The communication link is improved at four different design levels. The human body undergoes constant movement, which causes the transducer's orientation and the communication channel to change over time. Therefore, this work proposes to use small mm-scale, omnidirectional transducers. As a challenge, however, the channel multipath increases, and the received signal-to-noise ratio (SNR) worsens.

No channel models exist for ultrasound in-body communication with small-scale transducers. Therefore, this work empirically characterizes such in-body channels by realistically mimicking these with scattering gelatin phantoms and ex-vivo biomedical tissue. The obtained channel path loss is 55-105dB, including the acoustical transducer losses, and the multipath delay spread falls within a 100-150us duration. Further, a small dynamic channel test case on an anesthetized pig yields a channel time stability of 30ms.

Upon these channel model characterizations, the communication modem is optimized for high data rate and energy-efficient communication. The multipath channel impairments are addressed by a custom, multipath-tolerant orthogonal frequency division multiplexing (OFDM) modulation scheme optimized for the hardware and energy efficiency needs of a miniaturized implant. Simulations show an x10 improvement in link reliability on severe multipath channels over a non-equalized quadrature amplitude modulation scheme. Ex-vivo measurements on biomedical tissue demonstrate a data rate of 340kbps at a bit-error rate (BER) below 1e-4 and implantation depth of 10cm.

Finally, device miniaturization is obtained through the demonstration of a compact, end-to-end transceiver integrated circuit, featuring the first ultrasound in-body OFDM transceiver ASIC. This ASIC is made reconfigurable through a 50dB dynamic range Tx/Rx chain, a flexible modulation depth and bandwidth, and reprogrammable baseband processing, all to address the diverse in-body channel needs. Ex-vivo communication experiments on a 14cm long biomedical tissue achieve a 470kbps data rate with a BER of 3e-4. The experiments are not limited by the receiver SNR but rather by channel multipath, stating that a multipath-tolerant modem, such as OFDM, is a fundamental enabler to achieve reliable, low-energy, miniaturized in-body communication.

Date:3 Oct 2016 →  21 Aug 2023
Keywords:ultrasound communication, in-body communication, body area networks, OFDM modulation
Disciplines:Nanotechnology, Design theories and methods
Project type:PhD project