Diffuse Optical Imaging and Spectroscopy of the Human Breast for Quantitative Oximetry with Depth Resolution.
Yu, Yang.
2011
-
Abstract: Near-infrared
spectral imaging for breast cancer diagnostics and monitoring has been a hot research
topic for the past decade. Here we present instrumentation for diffuse optical imaging
of breast tissue with tandem scan of a single source-detector pair with broadband light
in transmission geometry for tissue oximetry. The efforts to develop the continuous-wave
(CW) domain instrument ... read morehave been described, and a frequency-domain (FD) system is also
used to measure the bulk tissue optical properties and the breast thickness
distribution. We also describe the efforts to improve the data processing codes in the
2D spatial domain for better noise suppression, contrast enhancement, and spectral
analysis. We developed a paired-wavelength approach, which is based on finding pairs of
wavelength that feature the same optical contrast, to quantify the tissue oxygenation
for the absorption structures detected in the 2D structural image. A total of eighteen
subjects, two of whom were bearing breast cancer on their right breasts, were measured
with this hybrid CW/FD instrument and processed with the improved algorithms. We
obtained an average tissue oxygenation value of 87% ± 6% from the healthy breasts,
significantly higher than that measured in the diseased breasts (69% ± 14%) (p <
0.01). For the two diseased breasts, the tumor areas bear hypoxia signatures versus the
remainder of the breast, with oxygenation values of 49 ± 11% (diseased region) vs.
61 ± 16% (healthy regions) for the breast with invasive ductal carcinoma, and 58
± 8% (diseased region) vs 77 ± 11% (healthy regions) for ductal carcinoma in
situ. Our subjects came from various ethnical/racial backgrounds, and two-thirds of our
subjects were less than thirty years old, indicating a potential to apply the optical
mammography to a broad population. The second part of this thesis covers the topic of
depth discrimination, which is lacking with our single source-detector scan system.
Based on an off-axis detection method, we incorporated an additional detector to acquire
a second set of image independently. We then proposed an inner-product approach to
associate absorption structures detected in the on-axis image with those detected in the
off-axis image. The spatial coordinate difference for the same structure between the two
images is directly related to the depth of the corresponding structure, and the
monotonic dependence can be quantified by perturbation theory of the diffusion equation.
A preliminary phantom study shows good agreement between the measured and the actual
depth of embedded structures, and human measurements show the capability to assign a
depth coordinate to the more complex absorption structures inside the
breast.
Thesis (Ph.D.)--Tufts University, 2011.
Submitted to the Dept. of Biomedical Engineering.
Advisor: Sergio Fantini.
Committee: Fiorenzo Omenetto, Donna Slonim, and Brian Pogue.
Keywords: Biomedical Engineering, Engineering, and Medical Imaging and Radiology.read less - ID:
- gb19fh92k
- Component ID:
- tufts:21062
- To Cite:
- TARC Citation Guide EndNote
