Abstract

Optical image reconstruction for biomedical imaging and diagnostics is an inverse problem which requires regularization to stabilize the inverse process. Two essential elements in image reconstruction is prior information and noise. We clarify their different roles in reconstruction by adopting a statistical interpretation of inversion which results in a generalized Tikhonov regularization formalism. Reconstruction for a slab from the generalized Tikhonov regularization is presented.

© 2002 Optical Society of America