KEYWORDS: Breast, 3D modeling, Tumors, Monte Carlo methods, Photon transport, Tissues, Tissue optics, Scattering, Magnetic resonance imaging, Absorption
The scattering and absorption properties of human breast are very important for cancer diagnosis based on diffuse
optical tomography (DOT). In this study, the dynamics of photon migration in three-dimensional human breast model
with various source-detector separations is simulated based on a Monte Carlo algorithm. The three-dimensional human
breast structure is obtained from in vivo MRI image. The breast model consists of skin, fatty tissue, glandular tissue,
sternum and ribcage. The backscattered diffuse photons from each layer in breast are recorded by marking the deepest
layer which every photon can reach. The experimental results indicate that the re-emitted photons contain more
information from deep tissue regions with the source-detector separations because of the strong dependence to the
resolution and sensitivity in DOT imaging. The geometric position of the source-detector separations were optimized in
this study. The different sizes of breast tumor were modeled to analysis of optical image characterizations. Finally, the
tumor images from different deep information were obtained with temporal profiles.
Breast cancer has been one of the leading causes of cancer deaths for females in the developed countries, including
the US. While early detection of breast cancer is essential for the reduction of death rate, there may be already more
than 107 cells in a breast cancer when it can be observed by X-ray mammogram. As contrast, the passive IR spectrogram
proposed by Szu et al. was shown to be promising in detecting the breast cancer several months ahead of mammogram.
With the energy readings from two IR cameras, one middle wavelength IR (MIR, 3 - 5μm) and one long wavelength IR
(LIR, 8 - 12μm), the IR spectrogram may be computed by using the blind source separation (BSS) algorithms developed
by Szu et al., which reveals the probability of being a cancer point on the breast surface. Two important tasks are
involved in computing the IR spectrogram. One is an accurate estimate of the ground state energy in the Helmholtz free
energy, H = E-T0S. The other is a correct pair-up of the points on the MIR and LIR images for a better estimation of
IR spectrogram. To minimize the probability of making an erroneous estimate of the ground state energy inherent in the
deterministic neighborhood-based BSS algorithm, a spatiotemporal approach is proposed in this paper. It takes into
account not only the neighborhood information but also the temporal information in determining the probability of being
a cancer point. Furthermore, a new sub-pixel super-resolution registration algorithm incorporating a third energy
dimension is proposed to establish better correspondences between the points in the MIR and LIR images. Phantom
study has confirmed that sub-pixel registration can be achieved by the proposed registration method. Human subject
study further shows that the breast cancer may be detected by the proposed spatiotemporal approach via
cross-referencing the IR spectrograms computed from the multiple pairs of MIR and LIR images taken at different times.
Bias field is a common phenomenon in a breast sonogram. Although artifacts caused by bias filed may carry important
information, e.g., shadowing behind a lesion, they are generally disturbing in the process of automatic boundary
delineation for sonographic breast lesions. This paper presents a new segmentation algorithm aiming to decompose the
region of interest (ROI) into prominent components while estimating the bias field in the ROI. A prominent component is
a contiguous region with a visually perceivable boundary, which might be a noise, an artifact, a substructure of a tissue or
a part of breast lesion. The prominent components may be used as the basic constructs for a higher level segmentation
algorithm to identify the lesion boundary.
The bias field in an ROI is modeled as a spatially-variant Gaussian distribution with a constant variance and
spatially-variant means, which is a polynomial surface of order n. The true gray levels of the pixels in a prominent
component are assumed to be Gaussian-distributed. The proposed algorithm is formulated as an EM-algorithm composed
of two major steps. In the E-step, the ROI is decomposed into prominent components using a new fuzzy cell-competition
algorithm based on the bias field and model parameters estimated in the previous M-step. In the M-step, the bias field
and model parameters are estimated based on the prominent components derived in the E-step using a least squared
approach. The results show that the effect of bias field on segmentation has been reduced and better segmentation results
have been attained.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.