Andy Kin On Wong1, Zamir Merali2, Jonathan D Adachi3. 1. Department of Medicine, McMaster University, Faculty of Health Sciences, Hamilton, Ontario, Canada; University Health Network, Osteoporosis Program, Toronto General Research Institute, Toronto, Ontario, Canada. Electronic address: wongko@mcmaster.ca. 2. University Of Toronto, Faculty of Medicine, Toronto, Ontario, Canada. 3. Department of Medicine, McMaster University, Faculty of Health Sciences, Hamilton, Ontario, Canada.
Abstract
OBJECTIVE: The purpose of this study was to develop a skeletal muscle mimic phantom compatible with quantitative computed tomography (QCT) and magnetic resonance imaging, yielding physiologically appropriate values. METHODS: Agar-based phantoms contained varying concentrations of CuCl2 and EDTA to adjust T2 relaxation time and phantom density concurrently. T2 relaxation times were quantified using a 4-mm single-slice fast spin echo sequence repeated for six serial echo times at 937-μm resolution. T2 relaxation maps were generated using the Levenberg-Marquardt equation. A peripheral QCT scanner measured linear attenuation coefficients of phantoms, which were converted to density (mg/cm3) values. Five 2.3 ± 0.5 mm thick slices were acquired at 15 mm/s scan speed and 500-μm resolution. Logarithmic or linear regression models were fitted to EDTA or CuCl2 versus density and T2 relaxation data. RESULTS: Density (D) was linearly dependent on CuCl2 (D = 0.27 [CuCl2] + 63.92, R2 = 0.84, P = 0.01) and invariant to EDTA. T2 relaxation time was related negatively to CuCl2 (T2 = -10.13 ln [CuCl2] + 66.70, R2 = 0.91, P < .01) and positively to EDTA (T2 = 5.72 ln [EDTA] + 54.47, R2 = 0.86, P < .01). Reproducibility within and between phantoms of the same compositions was acceptable (<5% error). Long-term stability was achieved for density but poorer for T2 relaxation time. CONCLUSIONS: This phantom optimization method provides a means for altering a soft tissue phantom suited for calibrating magnetic resonance imaging and QCT signals within values representative of muscle. Phantoms can be used during scans for calibrating magnetic resonance signals between and within individuals over time and can cross-calibrate different scanners.
OBJECTIVE: The purpose of this study was to develop a skeletal muscle mimic phantom compatible with quantitative computed tomography (QCT) and magnetic resonance imaging, yielding physiologically appropriate values. METHODS:Agar-based phantoms contained varying concentrations of CuCl2 and EDTA to adjust T2 relaxation time and phantom density concurrently. T2 relaxation times were quantified using a 4-mm single-slice fast spin echo sequence repeated for six serial echo times at 937-μm resolution. T2 relaxation maps were generated using the Levenberg-Marquardt equation. A peripheral QCT scanner measured linear attenuation coefficients of phantoms, which were converted to density (mg/cm3) values. Five 2.3 ± 0.5 mm thick slices were acquired at 15 mm/s scan speed and 500-μm resolution. Logarithmic or linear regression models were fitted to EDTA or CuCl2 versus density and T2 relaxation data. RESULTS: Density (D) was linearly dependent on CuCl2 (D = 0.27 [CuCl2] + 63.92, R2 = 0.84, P = 0.01) and invariant to EDTA. T2 relaxation time was related negatively to CuCl2 (T2 = -10.13 ln [CuCl2] + 66.70, R2 = 0.91, P < .01) and positively to EDTA (T2 = 5.72 ln [EDTA] + 54.47, R2 = 0.86, P < .01). Reproducibility within and between phantoms of the same compositions was acceptable (<5% error). Long-term stability was achieved for density but poorer for T2 relaxation time. CONCLUSIONS: This phantom optimization method provides a means for altering a soft tissue phantom suited for calibrating magnetic resonance imaging and QCT signals within values representative of muscle. Phantoms can be used during scans for calibrating magnetic resonance signals between and within individuals over time and can cross-calibrate different scanners.
Authors: Ainsley C J Smith; Justin J Tse; Tadiwa H Waungana; Kirsten N Bott; Michael T Kuczynski; Andrew S Michalski; Steven K Boyd; Sarah L Manske Journal: PLoS One Date: 2022-10-17 Impact factor: 3.752