RT Journal Article
SR Electronic
T1 Molecular Breast Tomosynthesis using a dual-detector, variable-angle slant hole system.
JF Journal of Nuclear Medicine
JO J Nucl Med
FD Society of Nuclear Medicine
SP 217
OP 217
VO 59
IS supplement 1
A1 David Gilland
A1 Benjamin Welch
A1 Seung Joon Lee
A1 BRIAN KROSS
A1 Andrew Weisenberger
YR 2018
UL http://jnm.snmjournals.org/content/59/supplement_1/217.abstract
AB 217Objectives: he objective of this work was to evaluate image quality obtained from a molecular breast tomosynthesis (MBT) system composed of two opposing detectors equipped with variable-angle, slant hole (VASH) collimators. MBT uses limited angle tomography and has demonstrated improved contrast-to-noise ratio over planar molecular breast imaging methods.1,2 The VASH collimator3 allows limited angle SPECT data to be acquired while the gamma camera remains stationary and close to the breast for optimal spatial resolution/sensitivity trade-off. The VASH collimator was constructed from a stack of 49 tungsten plates (0.265 mm plate thickness), each photo-etched with a 2D array of square holes (1.35 mm side length, 0.25 mm septal thickness). Aligning the plates creates a parallel hole collimator; shearing the plates to varying degrees creates a variable-angle, slant hole collimator. The shearing is precisely controlled by two pairs of stainless steel blocks that translate along opposite sides of the collimator. A phantom was built to simulate the Tc-99m distribution in molecular breast imaging and contained hot, spherical lesions in a warm background with a 20-to-1 concentration ratio. A range of lesion sizes from 3.9 mm to 7.8 mm diameter were used, and lesions were positioned throughout the 60 mm depth of the phantom. Data were acquired with the phantom resting on the collimator and again after flipping the phantom to simulate data from an opposing detector. For both phantom positions, the MBT data were acquired over a 50 degree angular range using 5 degree increments. Both high count level and clinical count level data were obtained. The MBT projection data were reconstructed using a ML-EM algorithm adapted for the limited angle geometry. The algorithm used 100 iterations followed by post-filtering using a Metz filter function. The reconstructed MBT images were compared to conventional planar images of the phantom at the opposing angles. The results showed that the reconstructed MBT images with the high count data allowed assessment of the spatial resolution in three dimensions. As expected, the spatial resolution in the depth dimension was degraded due to the limited angle nature of the data. In the clinical count density MBT images, the 6.2 mm lesions were clearly detectable at all depths and positions. In these images, the 4.9 mm lesions were also detectable, although noise structures at this size were present in the images with nearly equal contrast. The 3.9 mm lesions were not detectable. In conclusion, the proposed MBT system using two opposing gamma cameras equipped with VASH collimators demonstrated the ability to clearly detect lesions down to 6 mm size at all depths in the 60 mm phantom. The VASH collimator allows an MBT system design that is mechanically simple and maintains close proximity to the breast throughout the acquisition for optimal spatial resolution/sensitivity trade-off.