TY - JOUR
T1 - Magnetic resonance-guided near-infrared tomography of the breast
AU - Brooksby, Ben
AU - Jiang, Shudong
AU - Dehghani, Hamid
AU - Pogue, Brian W.
AU - Paulsen, Keith D.
AU - Kogel, Christine
AU - Doyley, Marvin
AU - Weaver, John B.
AU - Poplack, Steven P.
N1 - Copyright:
Copyright 2011 Elsevier B.V., All rights reserved.
PY - 2004/12
Y1 - 2004/12
N2 - The design and implementation of a multispectral, frequency-domain near infrared tomography system is outlined, which operates in a MRI magnet for utilization of MR-guided image reconstruction of tissue optical properties. Using long silica optical fiber bundles, measurements of light transmission through up to 12 cm of female breast tissue can be acquired simultaneously with MRI scans. The NIR system utilizes six optical wavelengths from 660 to 850 nm using intensity modulated diode lasers nominally working at 100 MHz. Photomultiplier tube detector gain levels are electronically controlled on a time scale of 200 ms, thereby allowing rapid switching of the source to locations around the tissue. There are no moving parts in the detection channels and for each source position, 15 PMTs operating in parallel allow sensitivity down to 0.5 pW/cm 2 at the tissue surface. Images of breast tissue optical absorption and reduced scattering coefficients are obtained using a Newton-type reconstruction algorithm to solve for an optimal solution using the measurement data. In medical imaging, it is beneficial to compare the same tissue volume as seen by a variety of modalities, and perhaps more importantly, there is the hypothesis that one imaging system which has high spatial resolution can be used to enhance the reconstruction of another system which has good contrast resolution. In this study we explore the synergistic benefits of a combined NIR-MRI data set, specifically the ways in which MRI (i.e., high spatial resolution) enhances NIR (i.e., high contrast resolution) image reconstruction. The design, calibration, and performance of the imaging system are described in the context of preliminary phantom tests and initial in vivo patient imaging. Co-registered MRI validates and improves optical property estimation in 2D tomographic image reconstructions when specialized algorithms are used.
AB - The design and implementation of a multispectral, frequency-domain near infrared tomography system is outlined, which operates in a MRI magnet for utilization of MR-guided image reconstruction of tissue optical properties. Using long silica optical fiber bundles, measurements of light transmission through up to 12 cm of female breast tissue can be acquired simultaneously with MRI scans. The NIR system utilizes six optical wavelengths from 660 to 850 nm using intensity modulated diode lasers nominally working at 100 MHz. Photomultiplier tube detector gain levels are electronically controlled on a time scale of 200 ms, thereby allowing rapid switching of the source to locations around the tissue. There are no moving parts in the detection channels and for each source position, 15 PMTs operating in parallel allow sensitivity down to 0.5 pW/cm 2 at the tissue surface. Images of breast tissue optical absorption and reduced scattering coefficients are obtained using a Newton-type reconstruction algorithm to solve for an optimal solution using the measurement data. In medical imaging, it is beneficial to compare the same tissue volume as seen by a variety of modalities, and perhaps more importantly, there is the hypothesis that one imaging system which has high spatial resolution can be used to enhance the reconstruction of another system which has good contrast resolution. In this study we explore the synergistic benefits of a combined NIR-MRI data set, specifically the ways in which MRI (i.e., high spatial resolution) enhances NIR (i.e., high contrast resolution) image reconstruction. The design, calibration, and performance of the imaging system are described in the context of preliminary phantom tests and initial in vivo patient imaging. Co-registered MRI validates and improves optical property estimation in 2D tomographic image reconstructions when specialized algorithms are used.
UR - http://www.scopus.com/inward/record.url?scp=16644369382&partnerID=8YFLogxK
U2 - 10.1063/1.1819634
DO - 10.1063/1.1819634
M3 - Article
AN - SCOPUS:16644369382
SN - 0034-6748
VL - 75
SP - 5262
EP - 5270
JO - Review of Scientific Instruments
JF - Review of Scientific Instruments
IS - 12
ER -