Assessing the spatial extent of breast tumor intrinsic optical contrast using ultrasound and diffuse optical spectroscopic imaging.

Abstract

Abstract Abstract #806 Background: Near infrared (NIR) light is sensitive to several important tissue components and has been used to detect, characterize, and monitor breast tumor functions. However, little is known about the relationship between anatomic and functional contrast derived from intrinsic optical signals. Diffuse Optical Spectroscopic Imaging (DOSI) is a non-invasive, bedside functional imaging technique that quantifies the concentrations and molecular states of tissue hemoglobin, water, and lipids. Pilot studies have shown that DOSI may be a useful tool for breast lesion detection/characterization and therapeutic monitoring.
 Materials and Methods: DOSI data were compared with theoretical computer simulations to estimate the nature and extent of optical contrast with respect to standard ultrasound imaging. DOSI reflectance measurements were obtained from 10 breast tumor patients, with both malignant (N=5) and benign (N=5) lesions. Patient average ages: 35+/-7.8 years vs. 44.3+/-14.7 years, lesion depth 10.9+/-5.5 mm vs. 15.8+/-5.3 mm, and maximum size 13.0+/-6.0 mm vs. 14.8+/-4.9 mm for benign and malignant, respectively. Two trial tumor models were simulated to model actual tumor optical properties (hemoglobin): (a) a discrete target constrained to ultrasound dimensions, and (b) a distributed target that spatially extended beyond ultrasound dimensions.
 Results: High optical contrast (i.e., the ratio target to background) was observed for both benign (1.53+/-0.17) and malignant (2.06+/-0.4) lesions, even for this younger population. While the discrete target model could not match experiment and simulation, a distributed model provided excellent agreement. The spatial extent of optical properties in breast tumor was significantly greater than anatomical dimensions reported by ultrasound imaging. The extent of the distribution was greater in the malignant lesions than in the benign ones. Further simulations also suggested that invasive breast tumors in this population with anatomical size 10 mm may still be detectable at depth 30 mm using DOSI in reflectance.
 Discussion: The spatial extension of tumor optical contrast complicates co-registration of optical functional images with anatomic images. The true optical contrast is significantly higher in breast cancer tumors, likely 4-5x than the hemoglobin levels in normal tissues. Breast tumor optical contrast is more realistically modeled as a spatially distributed target, which facilities a fast 3D imaging scheme in reflectance geometry. Results from these topographic reconstructions will be discussed. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 806.