Page 384 - Biomedical Engineering and Design Handbook Volume 2, Applications
P. 384
362 DIAGNOSTIC EQUIPMENT DESIGN
Likewise, there was insufficient evidence to recommend for or against MRI screening as an
adjunct to mammography in women having heterogeneously or extremely radiodense breast
parenchyma on mammography. The expert concensus opinion of the panel was that there was no
evidence to support MRI screening in women having a less than 15 percent lifetime risk of developing
breast carcinoma (Saslow et al., 2007).
12.4.4 Nuclear Imaging Methods
Nuclear imaging involves the injection of pharmaceutical compounds that have been labeled with
radioisotopes. The compounds are selected such that they couple to some sort of biological process
such as blood flow, metabolic activity, or enzyme production, or such that they tend to accumulate at
specific locations in the body, for example, binding to certain cell receptor sites. Thus the relative con-
centration of these radiotracers in various areas of the body gives information about the relative
degree to which these biological activities are occurring. Measurement of this concentration distribu-
tion therefore provides functional information very different from the structural information supplied
by modalities such as x-ray mammography and ultrasound. For this reason, nuclear medicine techniques
are being explored as adjunct imaging approaches to the structurally oriented x-ray mammography.
Nuclear medicine tracers being considered for breast imaging can conveniently be divided into
two groups: those emitting single gamma rays and those emitting positrons. In both cases, the con-
centration distribution in the breast is mapped by one or more imaging gamma detectors placed in
proximity to the patient. In the case of single gamma emitters, the gamma detectors are equipped
with some type of physical collimators whose function is to create a unique correlation between a
point on the detector surface and a line through the spatially distributed radioactivity source (e.g., the
breast). Physical collimation is not necessary in the case of positron-emitting isotopes because of the
near colinearity of the gamma ray pair produced by annihilation of the emitted positron with a nearby
electron. However, opposing gamma detectors and timing circuitry must be used to detect the two
gamma rays of a pair in coincidence. Below we discuss some of the unique challenges presented by
nuclear imaging of the breast, and describe some technical solutions currently being explored.
Scintimammography
Conventional Scintimammography. Nuclear medicine imaging of the breast using a tracer emitting
single gamma rays and an imaging gamma detector fitted with a collimator is referred to as breast
scintigraphy, scintimammography, or more contemporarily with the advent of dedicated imaging sys-
tems, molecular breast imaging or breast specific gamma imaging. These terms are typically applied
to planar imaging of the breast rather than tomographic imaging, known as single photon emission
computed tomography, or SPECT. Early scintimammography studies used conventional, large gamma
cameras and imaged the prone patient whose breasts hung pendant through holes in the table. Prone
positioning was favored over supine positioning because gravity acts to pull the breast tissue away
from the chest wall. Typically two lateral views were obtained, and one anterior-posterior view to aid
in medial-lateral tumor localization. The majority of clinical trials to date have employed one of two
99m Tc-labeled pharmaceuticals; sestamibi or tetrafosmin. Reported values for sensitivity and speci-
ficity for planar scintimammography performed under these conditions vary according to several fac-
tors, a principle one being the distribution of lesion sizes represented in the particular study. In a
three-center European trial, sensitivities of 26, 56, 95, and 97 percent were reported for category pT1a
(<0.5 cm), pT1b (0.5 to 1.0 cm), pT1c (1.0 to 2.0 cm), and pT2 (>2 cm) cancers, respectively (Scopinaro
et al., 1997). A clinical trial (134 women scheduled for open breast biopsy were enrolled) investigating
the use of prone 99m Tc-sestamibi scintimammography for pT1 tumors (4.7 percent pT1a, 46.7 percent
pT1b, and 48.6 percent pT1c) reported sensitivity, positive predictive value, negative predictive value
and accuracy of 81.3, 97.6, 55.6, and 83.6 percent, respectively (Lumachi et al., 2001). The correspond-
ing values for x-ray mammography were 83.2, 89.9, 48.6, and 79.1 percent. In a recent meta-analysis,
Hussain et al. reported scintimammography sensitivity and specificity of 85 and 84 percent, respectively
for 2424 single-site participants, and 85 and 83 percent, respectively for 3049 multicenter study partici-
pants (Hussain et al., 2006). These and other studies demonstrated that scintimammography provides