Abstract: 

The invention of computerised tomography (CT) was a major breakthrough, because it was the first technology which allowed to reconstruct a full 3d picture of the human body. Now, several decades later, this technology has been refined and produces images of highest quality. There are, however, serious drawbacks of this method. It is very expensive, it has a low throughput rate and it is not suitable for monitoring purposes. Hence there is a demand for an alternative, which complements the strengths and weaknesses of CT.

Speaker

Dr. Janosch Rieger

Research Area
Affiliation

Imperial College London

Date

Thu, 28/04/2016 - 11:05am to 11:55am

Venue

RC-4082, The Red Centre, UNSW

The invention of computerised tomography (CT) was a major breakthrough, because it was the first technology which allowed to reconstruct a full 3d picture of the human body. Now, several decades later, this technology has been refined and produces images of highest quality. There are, however, serious drawbacks of this method. It is very expensive, it has a low throughput rate and it is not suitable for monitoring purposes. Hence there is a demand for an alternative, which complements the strengths and weaknesses of CT.

Electrical impedance tomography is a very modern, non-invasive and budget-priced technology for the detection of inhomogeneities, which is about to capture the medical technology market. First devices for the detection of tumors and real-time pulmonary function diagnostics have been built and are now available for purchase. Unfortunately, the reconstruction of an inhomogeneity from data is a difficult mathematical problem, that has not yet been solved in a satisfactory way, so the precision and reliability of the currently available devices is far from optimal. 

I will present a new mathematical approach to electrical impedance tomography, which I have developed in a collaboration with Bastian von Harrach. It builds on the concept of the convex source support from scattering theory, results from inverse problems and a novel technology for numerical optimisation in the space of the convex and compact subsets of R^d, and the results of first numerical tests are very promising.