Mammography is the most common technology used to characterize breast abnormalities in the diagnostic setting. Although mammography is a valuable screening and diagnostic technique, it has some significant limitations. Mammography has variable sensitivity: it is less sensitive in detecting breast cancer in women with radiographically dense breasts. Screening mammography has a low positive predictive value. There is also considerable variability in the radiologic interpretation of mammograms. Hence, there is considerable rationale to pursue alternative imaging methods for breast cancer detection. The principal hypothesis of Project IV is that frequency domain modulation methods for near infrared (NIR) optical spectroscopy of tissue coupled with model-based image formation from tomographically-measured optical data provides a functional imaging methodology with clinical relevance to the detection and diagnosis of breast cancer. This imaging method uses fundamentally new approaches for finite element based computed tomography reconstruction, along with new techniques for diffuse optical spectroscopy projection measurement through tissue.
The specific aims i nclude (1) constructing a parallelized 48-unit detector array for fast near- infrared (NIR) data acquisition (2) developing a new mechanical interface for clinical NIR breast scanning providing adequate optical fiber-tissue coupling and suitable precision translation for multi-slice image acquisition, (3) developing a compact diode-laser-driven spectroscopic imaging system covering the optical wavelengths between 600 and 850 nm which is sufficient to quantitatively image absorption coefficients in vivo and evaluate its functional imaging capabilities in terms of hemoglobin concentration image absorption coefficients in vivo and evaluate its functional imaging capabilities in terms of hemoglobin concentration, oxygen saturation, water and lipid concentrations, (4) advancing, in conjunction with the Computational Core, model-based image reconstruction technology for NIR absorption by investigation multi-component objection functions, adaptive spatial regularization, adaptive dual meshing, three-dimensional methods, parallel processing and multi-spectral modulation strategies for increasing property sensitivity, and (5) evaluating, in conjunction with the Clinical Core, the clinical potential of NIR breast imaging in a trial involving patients with normal and abnormal mammograms, with the goal of defining the available contrast for optical imaging and the pathobiologic basis for this contrast. While there is still no absolute proof that optical imaging will enhance the diagnosis or management of breast disease, much of the preliminary data indicates there is a difference in tissue composition (due to hemoglobin volume, much of the preliminary data indicates that there is a difference in tissue composition (due to hemoglobin volume, oxygenation, water content, calcification, etc.) which can be discriminated through the use of near-infrared spectroscopy; hence, the technique holds promise of exploiting new contrast mechanisms for breast cancer detection, which offers the possibility of adding new functional information to diagnostic decision-making.
Showing the most recent 10 out of 129 publications
