Resolved object imaging and localization with the use of a backpropagation algorithm.

Optical diffusion tomography is an emerging technology that generates images of objects imbedded in turbid media using scattered light. To date, however, most demonstrations of this technology use a sphere or a collection of spheres as the imbedded object. Here we use a backpropagation algorithm and a planar geometry to reconstruct images of resolved objects (airplane models) imbedded in tissue phantoms. In addition, we show that we can locate the resolved objects in three dimensions in the turbid medium using only a single planar view. The imaging system uses diffuse photon density waves produced using kilohertz modulation (that is, essentially dc illumination).