Summary

This work is on developing numerical reconstruction methods for free-space fluorescence diffuse optical tomography, using efficient regularization policies.

Introduction

The goal in fluorescence diffuse optical tomography (FDOT) is to reconstruct the distribution of fluorescent markers inside tissue based on boundary light measurements. The challenge is to invert the diffuse nature of light in biological tissue. This a non-invasive imaging technique applicable, for instance, to cancer detection or small animal imaging in drug development.

Main Contribution

We have developed an integrated framework based of the finite-elements method which combines a forward model and a reconstruction scheme. We are investigating sparsity-promoting regularization techniques, which yield better reconstruction quality than standard linear techniques. Our research also covers the improvement of the data acquisition scheme, and the incorporation of a priori knowledge from other imaging modalities.