Effect of random noise on a bio-optical imaging scheme

Simulations for detecting objects inside of biological materials by using laser light are presented. Our method uses the weighted shadow patterns of objects at fixed locations to ``mask'' one section of the detection region, so that objects in the unmasked region can be located by scanning through various locations and minimizing $\chi^2$. It is shown that when random noise is introduced, the method is very robust when only a single object is present in the medium but that detection of multiple objects can be very sensitive to noise.