Locating inhomogeneities in tissue by using the most probable diffuse path of light.

Jing Bai,Tianxin Gao,Kui Ying,Nanguang Chen
DOI: https://doi.org/10.1117/1.1896968
IF: 3.758
2005-01-01
Journal of Biomedical Optics
Abstract:Characterization of human tissue using near-IR (NIR) light is becoming increasingly popular. The light signal transmitted from the tissue contains information concerning inhomogeneities in tissue, such as size, position, and pathological states (benign or malignant). We discuss the most probable diffuse path (MPDP) related to frequency-domain diffuse photon density waves (DPDWs) propagating inside turbid media. We find that for a medium of finite size, the existence of boundaries between tissue and nonscattering media would have considerable impact on the path shape. It is also demonstrated that such paths can be used to obtain higher accuracy in localizing absorbers embedded in a homogeneous background. Based on the proposed MPDP, a new method for 3-D localization of hetero-geneities in turbid media is proposed, which is validated by experiments using Intralipid and pork fat. The experiments are performed with an NIR breast cancer detection system designed and assembled in our lab, using 780-nm NIR light. In Intralipid, when the size of a single absorber is less than 1 cm, the localization error is about 2 mm. The results from pork fat are also acceptable. (c) 2005 Society of PhotoOptical Instrumentation Engineers.
What problem does this paper attempt to address?