Signal processing offers real opportunities to improve diagnosis and management of clinical medical conditions. Signal processing applications have been slow to be developed relative to research and development in drugs, biomedical instrumentation and basic science. The gap between the huge and productive investment in basic science research and what we now know in basic science and how this knowledge can be used to impact on clinical medical outcomes through the translational research process is now achieving international focus. Signal processing will be important in this new wave of translational research.

I discuss several examples from my own research, where we have needed and successfully used signal processing in the translational research setting. These examples will focus on signal processing of the electroencephalogram (EEG). EEG abnormalities occur in the newborn baby and need to be identified. An automated method of detecting abnormalities is also essential in the investigation of new treatments in the context of randomized clinical trials. A second example is a project we are presently engaged in in which we are developing biomedical instrumentation through to a target of preventing death of the fetus in utero (stillbirth).

Further applications and opportunities for signal processing to improve health outcomes will be presented and discussed.

Astronomy is one of the oldest fields of science and yet it is one of the most popular. Starting with the earliest human beings, the celestial bodies have not ceased to fascinate mankind and the work of Galileo Galilei in developing a telescope and work of various scientists after him such as Gauss on the analysis of time-series data have opened paths to the accelerated the development of this field of science. The efforts in the development of observational and deductive astronomy led to the foundations of many important concepts such as least squares, curve fitting, minimax theory, nonlinear spectrum analysis and so on.

These very days we are in, the field is living again a very exicting period with the arrival of measurement results of  a number of satellite missions. In particular, the Planck and Herschel satellites of European Space Agency have been launched on 14 May 2009 to much excitement of the world scientific community and the world press. Another satellite mission, WMAP of NASA is continuing sending us multichannel measurements in microwave channels. Planck and WMAP missions have the objective of making radiation maps of the full sky to uncover high resolution images of the cosmic microwave background radiation while Herschel will study the formation of galaxies in the early universe and their subsequent evolution. Since the discovery of the CMB radiation in 1965 by Penzias and Wilson, a number of missions have been planned to measure the CMB. CMB radiation is of interest from a number of aspects:  it is the most important evidence for the hot big-bang model; it provides us a picture of the universe in its very early moments; the anisotropies in it provide the seed map of the universe of today; it provides us information about the fundamental constants and hence the future of our universe. Unfortunately, the task of measuring CMB is not an easy one, since radiation measurements of the sky contain contributions from various sources from our galaxy such as synchrotron, galactic dust and free-free emission as well as extragalactic radio sources. The challenge of recovering CMB from these multichannel observations has led to the development of various new methodology for source separation over the last decade. The need to detect galaxies which depict themselves as sparse compact sources provides new challenges in detection theory.

Satellite missions and telescopes provide vast amounts of data which need to be sifted with data mining techniques to recover features of interest. In some cases, the data are redundant due to sky scanning strategies and certain signals such as point sources demonstrate sparsity. This nature of the data provoked research into tailored compression schemes including compressive sensing. The problem of weak lensing, the process in which light from distant galaxies is bent by the gravity of intervening mass in the universe, the imperfect imaging by telescopes due to finite point spread functions and device noise require novel image inversion methods with realistic signal and noise models.

These and many other problems make astronomy still a field which inspires the development of signal processing methodology. In this talk, we will present some of the current signal processing challenges in astronomy and indicate open problems and future research directions.

Signal processing is quite intensive when dealing with radar sensors. In our presentation, we will not, of course, detail this processing. Instead, we will present the main characteristics of the data acquisition, and data specifics and statistics. Indeed, one must understand the acquisition process in order to extract the useful information from the data. The more versatile tools are actually active sensors either airborne or spaceborne. I will focus here more precisely on the Synthetic Aperture Radar (SAR). From a signal processing perspective, radar data and images are a very interesting domain of research. Indeed, they present non-gaussian, non-stationary and often non-linear characteristics, thus raising the challenge of their interpretation.

The fact that the electromagnetic wave is in interaction with the observed media adds to the difficulty as the result is a convolution process. Also, the “radar eye” is far from being comparable to the optical equivalent. Even though the end result is often an image, the interpretation must not forget that the radar observe the scene through an antenna beam with its own geometry. I will present the ability of spaceborne radars for imaging ocean surface features and some of the very actual results will be shown through several illustrations. The imaging mechanism will be explained in detail and, again, a signal processing interpretation will be given. The signal processing tools for recovering the geophysical information will be described. I will in particular focus on 2D signal analysis tools for processing the data when they are available as “images”.

The aim of such means to watch the earth is to provide useful information on a global basis regardless of the weather conditions (especially clouds) and in daylight as well as at nighttime. The large spatial coverage of the observations, coupled with a quite good temporal revisit time allows us to reach a spatial structure in the framework of the monitoring of oceans and land for climate change purposes. Scientists will therefore be able to help in societal development and spatial planning at the regional scale, for a better warning for natural or man-made hazards, such as oil pollution, flood, etc. and in the spirit of territorial cohesion.