Validation of MRI Quantification of Liver Fat, Keeping CT Liver Attenuation Index (LAI) as Gold Standard

Abstract

Abstract: Background: Non – Alcoholic Fatty Liver Disease (NAFLD) is emerging as a considerable health problem in patients visiting gastroenterology clinics. It is of crucial importance to evaluate the extent of hepatic steatosis in potential candidates for living donor liver transplantation (LDLT) to ensure donor safety as well as optimum graft regeneration. Objective: To validate the MRI quantification of liver fat keeping CT liver attenuation index as gold standard. Materials and Methods: This cross-sectional study was carried out in the department of Diagnostic Radiology, Pakistan Kidney and Liver Institute and Research Centre from 10th October, 2022 to 10th December, 2022. We determined the sample size using WHO sample size calculator. The MR fat fraction sequence was acquired as a part of the obligatory MRCP in 70 potential liver donors who undertook CT abdomen. Liver Attenuation Index (LAI) and MR fat fraction were determined separately by two radiologists who were blinded to each other. LAI was calculated as: Mean liver attenuation – mean splenic attenuation. MRI fat fraction from seven areas of liver were taken and their mean calculated to determine the percentage of liver fat. SPSS version 20 was employed for statistical analysis and Pearson’s Correlation was applied. Results: Among the 70 donors 42 were males and 28 were females (M: F= 1.5: 1). The hepatic fat fraction values on MR were correlated with the liver attenuation index on CT using a two – tailed Pearson correlation test. The results showed a very strong negative correlation between the two; the lower the LAI, the higher the MR fat fraction (Pearson correlation coefficient r = -0.932, p<0.05). Conclusion: Strong correlation was found between MRI estimation of liver fat and CT LAI fat estimation. MRI is safer than CT as it does not involve ionizing radiation, is quicker to perform, and hence can be recommended as future method of choice.