Àá½Ã¸¸ ±â´Ù·Á ÁÖ¼¼¿ä. ·ÎµùÁßÀÔ´Ï´Ù.
KMID : 1149020210230010001
Journal of Korean Society of Computed Tomographic Technology
2021 Volume.23 No. 1 p.1 ~ p.10
Liver Fat Measurement with Quantitative CT and Correlation with MRI
Lim Bo-Hee

Lee Seong-Ju
Jeong Sang-Heon
Lee Min-Su
Bang Yong-Sik
Abstract
The prevalence of fatty liver has been continuously rising and is projected to become the most common indication for liver transplantation. Accumulation of fat in the liver exceeded 5% is dangerous for liver donors and recipients. Liver biopsy has been the reference standard for assessment of liver fat but it is invasive and MRI is only allowed for patients with no pacemaker and metallic implants. The purpose of this study was to validate the accuracy of quantitative CT in measurement of liver fat by using Dual Energy Computed Tomography(DECT) factors like Syngo.via Fat map Fat Fraction, DECT Pre image Mix 0.6 HU value and Virture Non Contrast(VNC) HU value. Sixty liver pre donors who underwent DECT and Magnetic Resonance Spectroscopy(MRS) were included. Linear correlation exists between MRS Fat Fraction and DECT factors yielding the following conversion formulas: Fat Fraction(%)= 0.9161¡¿Fat map Fat Fraction(%)+10.501, Fat Fraction(%)= -0.6435¡¿DECT Pre image Mix 0.6 HU value+45.851, Fat Fraction(%)= -0.6162¡¿VNC HU value+46.716. Statistical analysis was performed using SPSS 18.0 paired t-test. Fat Fraction calculated by conversion formula using Syngo.via Fat map Fat Fraction and MRS Fat Fraction has no significant difference(p=0.999), Fat Fraction calculated by conversion formula using DECT Pre image Mix 0.6 HU value has no significant difference(p=0.996), and also using VNC HU value has no significant difference(p=0.997). The most accurate Fat Fraction calculated by conversion formula was using DECT Pre image Mix 0.6 HU value (RMSE=2.70). The linear correlation between Fat Fraction derived from the conversion formulas and MRS Fat Fraction allows quantification of liver fat by means of DECT.
KEYWORD
Dual Energy Computed Tomography, Magnetic Resonance Spectroscopy, Fat Fraction
FullTexts / Linksout information
Listed journal information