Molecular breast imaging-A phantom study on the impact of collimator selection on the detection of sub-10 mm breast lesions

Carrie B Hruska, Michael K. O'Connor

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Molecular breast imaging (MBI) with a dedicated breast gamma camera system can detect small breast lesions with a sensitivity of >85%. The objective of this study was to determine the optimal collimation for tumor detection based on count densities obtained from clinical MBI studies. Seven collimators were evaluated using a cadmium zinc telluride detector designed for breast imaging. These included LEUHR, LEHR, LEGP, LEHS, and LEUHS collimators and two system-specific collimators-a long bore (LB) and general-purpose (GP) collimator with square holes matched to the detector elements. Collimators were evaluated using a breast phantom comprising a 20×20×20 cm plastic box containing 16 glass "tumors" with internal diameters ranging from 4-10 mm. Breast thickness was set to 6 cm and tumor depth was varied from 1-5 cm. The phantom and spheres were filled with water and Tc-99m to give a tumor to background (T/B) ratio varying from 3:1 to 35:1. Total counts acquired in each image simulated the range of count densities observed clinically. Counts acquired were adjusted to compensate for differences in collimator sensitivity. Tumor signal-to-noise ratio (SNR) was measured through ROI analysis. Images acquired at clinical count densities contained significant amounts of noise, especially at T/B ratios of 10:1 or less. Highest tumor SNR was obtained with the LEHS collimator for the 6, 8, and 9 mm tumors at depths of 1 and 3 cm. At a tumor depth of 5 cm, the highest SNR was obtained with either the matched GP or LEHS collimators for the 6-9 mm tumors. Low SNR was obtained with all collimators for the 4 mm tumors at 1 and 3 cm and no 4 mm tumors were visible at a depth of 5 cm. High sensitivity collimators may be better than high-resolution collimators for detecting tumors <1 cm in low count images of the breast, especially for tumors located within 1-4 cm of the collimator face, but proper collimator design to eliminate aliasing artifacts is important for pixilated systems.

Original languageEnglish (US)
Pages (from-to)250-254
Number of pages5
JournalNuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
Volume569
Issue number2 SPEC. ISS.
DOIs
StatePublished - Dec 20 2006

Fingerprint

collimators
breast
lesions
Tumors
tumors
Imaging techniques
Signal to noise ratio
signal to noise ratios
sensitivity
Detectors
zinc tellurides
cadmium tellurides
detectors
collimation
Cadmium
boxes
artifacts
Zinc
plastics
Cameras

Keywords

  • Breast tumor
  • Cadmium zinc telluride
  • Collimator
  • Molecular breast imaging

ASJC Scopus subject areas

  • Instrumentation
  • Nuclear and High Energy Physics

Cite this

@article{b03218bc7241429ab2b1ce7f42ae461d,
title = "Molecular breast imaging-A phantom study on the impact of collimator selection on the detection of sub-10 mm breast lesions",
abstract = "Molecular breast imaging (MBI) with a dedicated breast gamma camera system can detect small breast lesions with a sensitivity of >85{\%}. The objective of this study was to determine the optimal collimation for tumor detection based on count densities obtained from clinical MBI studies. Seven collimators were evaluated using a cadmium zinc telluride detector designed for breast imaging. These included LEUHR, LEHR, LEGP, LEHS, and LEUHS collimators and two system-specific collimators-a long bore (LB) and general-purpose (GP) collimator with square holes matched to the detector elements. Collimators were evaluated using a breast phantom comprising a 20×20×20 cm plastic box containing 16 glass {"}tumors{"} with internal diameters ranging from 4-10 mm. Breast thickness was set to 6 cm and tumor depth was varied from 1-5 cm. The phantom and spheres were filled with water and Tc-99m to give a tumor to background (T/B) ratio varying from 3:1 to 35:1. Total counts acquired in each image simulated the range of count densities observed clinically. Counts acquired were adjusted to compensate for differences in collimator sensitivity. Tumor signal-to-noise ratio (SNR) was measured through ROI analysis. Images acquired at clinical count densities contained significant amounts of noise, especially at T/B ratios of 10:1 or less. Highest tumor SNR was obtained with the LEHS collimator for the 6, 8, and 9 mm tumors at depths of 1 and 3 cm. At a tumor depth of 5 cm, the highest SNR was obtained with either the matched GP or LEHS collimators for the 6-9 mm tumors. Low SNR was obtained with all collimators for the 4 mm tumors at 1 and 3 cm and no 4 mm tumors were visible at a depth of 5 cm. High sensitivity collimators may be better than high-resolution collimators for detecting tumors <1 cm in low count images of the breast, especially for tumors located within 1-4 cm of the collimator face, but proper collimator design to eliminate aliasing artifacts is important for pixilated systems.",
keywords = "Breast tumor, Cadmium zinc telluride, Collimator, Molecular breast imaging",
author = "Hruska, {Carrie B} and O'Connor, {Michael K.}",
year = "2006",
month = "12",
day = "20",
doi = "10.1016/j.nima.2006.08.034",
language = "English (US)",
volume = "569",
pages = "250--254",
journal = "Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment",
issn = "0168-9002",
publisher = "Elsevier",
number = "2 SPEC. ISS.",

}

TY - JOUR

T1 - Molecular breast imaging-A phantom study on the impact of collimator selection on the detection of sub-10 mm breast lesions

AU - Hruska, Carrie B

AU - O'Connor, Michael K.

PY - 2006/12/20

Y1 - 2006/12/20

N2 - Molecular breast imaging (MBI) with a dedicated breast gamma camera system can detect small breast lesions with a sensitivity of >85%. The objective of this study was to determine the optimal collimation for tumor detection based on count densities obtained from clinical MBI studies. Seven collimators were evaluated using a cadmium zinc telluride detector designed for breast imaging. These included LEUHR, LEHR, LEGP, LEHS, and LEUHS collimators and two system-specific collimators-a long bore (LB) and general-purpose (GP) collimator with square holes matched to the detector elements. Collimators were evaluated using a breast phantom comprising a 20×20×20 cm plastic box containing 16 glass "tumors" with internal diameters ranging from 4-10 mm. Breast thickness was set to 6 cm and tumor depth was varied from 1-5 cm. The phantom and spheres were filled with water and Tc-99m to give a tumor to background (T/B) ratio varying from 3:1 to 35:1. Total counts acquired in each image simulated the range of count densities observed clinically. Counts acquired were adjusted to compensate for differences in collimator sensitivity. Tumor signal-to-noise ratio (SNR) was measured through ROI analysis. Images acquired at clinical count densities contained significant amounts of noise, especially at T/B ratios of 10:1 or less. Highest tumor SNR was obtained with the LEHS collimator for the 6, 8, and 9 mm tumors at depths of 1 and 3 cm. At a tumor depth of 5 cm, the highest SNR was obtained with either the matched GP or LEHS collimators for the 6-9 mm tumors. Low SNR was obtained with all collimators for the 4 mm tumors at 1 and 3 cm and no 4 mm tumors were visible at a depth of 5 cm. High sensitivity collimators may be better than high-resolution collimators for detecting tumors <1 cm in low count images of the breast, especially for tumors located within 1-4 cm of the collimator face, but proper collimator design to eliminate aliasing artifacts is important for pixilated systems.

AB - Molecular breast imaging (MBI) with a dedicated breast gamma camera system can detect small breast lesions with a sensitivity of >85%. The objective of this study was to determine the optimal collimation for tumor detection based on count densities obtained from clinical MBI studies. Seven collimators were evaluated using a cadmium zinc telluride detector designed for breast imaging. These included LEUHR, LEHR, LEGP, LEHS, and LEUHS collimators and two system-specific collimators-a long bore (LB) and general-purpose (GP) collimator with square holes matched to the detector elements. Collimators were evaluated using a breast phantom comprising a 20×20×20 cm plastic box containing 16 glass "tumors" with internal diameters ranging from 4-10 mm. Breast thickness was set to 6 cm and tumor depth was varied from 1-5 cm. The phantom and spheres were filled with water and Tc-99m to give a tumor to background (T/B) ratio varying from 3:1 to 35:1. Total counts acquired in each image simulated the range of count densities observed clinically. Counts acquired were adjusted to compensate for differences in collimator sensitivity. Tumor signal-to-noise ratio (SNR) was measured through ROI analysis. Images acquired at clinical count densities contained significant amounts of noise, especially at T/B ratios of 10:1 or less. Highest tumor SNR was obtained with the LEHS collimator for the 6, 8, and 9 mm tumors at depths of 1 and 3 cm. At a tumor depth of 5 cm, the highest SNR was obtained with either the matched GP or LEHS collimators for the 6-9 mm tumors. Low SNR was obtained with all collimators for the 4 mm tumors at 1 and 3 cm and no 4 mm tumors were visible at a depth of 5 cm. High sensitivity collimators may be better than high-resolution collimators for detecting tumors <1 cm in low count images of the breast, especially for tumors located within 1-4 cm of the collimator face, but proper collimator design to eliminate aliasing artifacts is important for pixilated systems.

KW - Breast tumor

KW - Cadmium zinc telluride

KW - Collimator

KW - Molecular breast imaging

UR - http://www.scopus.com/inward/record.url?scp=33751424827&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33751424827&partnerID=8YFLogxK

U2 - 10.1016/j.nima.2006.08.034

DO - 10.1016/j.nima.2006.08.034

M3 - Article

AN - SCOPUS:33751424827

VL - 569

SP - 250

EP - 254

JO - Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment

JF - Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment

SN - 0168-9002

IS - 2 SPEC. ISS.

ER -