Latest research results for 2008

• Enhancement and analysis of biologically produced signals using empirical mode decomposition and advanced spectrogram techniques [PDF]
• Frequency estimation and model estimation of bat echolocation calls [PDF]
• Motion compensated complementary coding for Medical ultrasound [PDF]
• Receiver characteristics consideration in statistical spectral estimation in microbubble echo signals [PDF]

Research results for 2007

• Binary coded signals for medical ultrasound [PDF]
• Estimation and detection of echo signals from
  ultrasound contrast microbubbles [PDF]
• Frequency tracking for bat echolocation calls [PDF]
• Signal analysis using the empirical mode
  decomposition [PDF]

The lead group for signal processing is the Institute for Digital Communications at the University of Edinburgh. A number of algorithms are under investigation for the processing of signals generated by BIAS researchers.
Several bat species exhibit remarkable performance in detecting targets in highly cluttered environments. For example, they are able to chase their prey in vegetation but do not collide with the branches and leaves. This achievement indicates that bats have the ability to decompose the returned acoustic signals (consisting of highly overlapping echoes) to the number of its constituent parts and process them adequately. Empirical mode decomposition (EMD) performs tasks that resemble the ones performed by bats. More specifically, EMD is aimed to decompose a signal to a number of amplitude and frequency modulated (AM/FM) zero mean components. It is hoped that EMD will produce less distortion than other processing methods and lead to accurate estimates of instantaneous frequency and phase.

Bat echolocation signals appear to follow certain structure, with certain types of signals adapted to certain environments. So far, only coarse signal structure has been observed in the signals using classical methods such as the short-time Fourier transform, or wavelet. A robust algorithm for frequency tracking is under development that should highlight finer detail in these calls, giving some insight into (echolocation) signal design and allow structural detail from bat echolocation signals to be extracted.

Algorithms are also being developed for the estimation and detection of echo signals from ultrasound contrast microbubbles. The first step is to find out the characteristics of echo signals from a solid-sphere and those from ultrasound contrast microbubbles. Some of the responses from microbubbles have one single pulse in the signal whereas other responses are composed of several pulses. The most important responses from the microbubble are its frequency components in the frequency domain and the pulse location in the time domain.