(Tuesday, 7/16/2019) 1:15 PM - 1:45 PM
Room: Exhibit Hall | Forum 9
Purpose: The accurate fat fraction is important information in classification of diseases in computed tomography (CT). However, the range of Hounsfield unit (HU) which is normalized value of attenuation coefficient in CT image for fat is vary as -50 to -150 HU. In this study, the feasibility of fat fraction based segmentation in CT by thresholding the fat range of HU based on surfactant-free microemulsion was investigated.
Methods: The feasibility of fat fraction based segmentation were studied first with digital phantom in the GATE simulation under the various tube voltages, 40, 50, 60, 70, and 100 kVp. The simulated fat fraction of microemulsion were 0 %, 25 %, 50 %, 75 %, and 100 %. The surfactant-free microemulsion of various fat fractions in the physical phantom were fabricated by high-intensity focused ultrasonic emulsification technique. All microemulsions in the various fat fraction were well-characterized and monitored for stability and accuracy. The physical phantom images were acquired with clinical CT system at 100 kVp. The segmentation ability was evaluated by normalized peak-to-valley ratio (NPR) which indicate resolving power between peaks in the image histogram.
Results: The NPR of fat fraction 0 to 100 % were 50 %, 59 %, 76 %, 85 %, and 81 % in digital phantom, respectively. For the microemulsion in physical phantom, peak HU values for each fat fraction 0 to 100 % were -5, -30, -70, -110, and -140 HU, respectively. The mean peak-to-peak reduction was about 34.75 HU. The NPR was 47 %, 45 %, 94 %, 38 %, and 32 %, respectively. The peaks and valleys among the fat fraction were clearly distinguished in the simulation and experimental study.
Conclusion: The results show that the well characterized surfactant-free microemulsion has the potential to be used for accurate fat fraction quantification of disease.
Funding Support, Disclosures, and Conflict of Interest: This work was supported in part by NRF-2018M3A9H6081482 and the KRISS grant 19011054.
Keywords
CT, Phantoms, Quantitative Imaging
Taxonomy
IM- CT: Quantitative imaging/analysis
Contact Email
