Abstract

Measurements of diffuse near-infrared light are used to tomographically image hemoglobin in tissue. A diffusion theory finite element solution is matches theoretical predictions to these measurements, allowing computed images of absorption and scattering coefficients to be recovered from the projection data by iteratively solving a regularized matrix equation which using the Newton-Raphson method. The diffuse measurements are recorded at wavelengths between 660 nm and 850 nm with automated data acqusition, using intensity-modulated light at 100 MHz. This prototype system is being tested on breast cancer patients to determine the utility of hemoglobin morphology for tumor diagnosis. Hemoglobin concentration and oxygen saturation images are derived from reconstructed absorption coefficient images at multiple wavelengths through a multi-parameter concentration fit at each point in the image. The chromophores assumed to be present in breast tissue are water, lipids, oxy- and deoxy-hemoglobin. This type of data acquisition system allows multi-spectral imaging which may provide the information needed to accurately recover the hemoglobin concentration and oxygen saturation images of the breast.

© 2000 Optical Society of America