1: \begin{abstract}

2: A numerical method for an inverse problem for an elliptic equation

3: with the running source at multiple positions is presented. This

4: algorithm does not rely on a good first guess for the solution. The

5: so-called \textquotedblleft approximate global convergence" property

6: of this method is shown here. The performance of the algorithm is

7: verified on real data for Diffusion Optical Tomography. Direct

8: applications are in near-infrared laser imaging technology for

9: stroke detection in brains of small animals.\bigskip

10: 

11: \begin{keywords}Approximate Global Convergence Property; Inverse Problem; Diffusion

12: Optical Tomography; Real Data

13: \end{keywords}

14: \begin{classcode}65B21; 65D10; 65F10\end{classcode}\bigskip

15: 

16: \end{abstract}