Output details

Approximation errors and model reduction in three-dimensional diffuse optical tomography

<27> When matching computational models to experimental data it is usually impossible to account for all systematic and physical differences between a model and reality. Used in image reconstruction, a deficient model leads to bias and increased noise in the solution. We address this directly by evaluating the statistical distribution of the errors between the model and the data, demonstrating on experimental optical tomographic image reconstruction. We reduce the computational time of the reconstruction from several hours to a few minutes opening the door to real-time reconstruction. We have published over 10 follow-on papers and gave invited talk MSRI Berkely(2010).