Fatty Liver Grading Using Computed Tomography Hounsfield Unit Values and Correlating With Ultrasonography Grading

Introduction Fatty liver disease (FLD), characterized by excessive fat accumulation in hepatocytes, is a growing public health concern linked to obesity, diabetes mellitus, and metabolic syndrome. Accurate assessment of FLD severity is essential for early diagnosis, monitoring disease progression, and implementing appropriate therapeutic interventions. This study aimed to compare the grading of FLD using computed tomography (CT) Hounsfield unit (HU) values with ultrasonography (USG) grading to determine the correlation and diagnostic accuracy between these two imaging modalities. Methods A cross-sectional observational study was conducted at Sri Devaraj Urs Medical College, Tamaka, Kolar, involving 110 adult participants aged 18 years and older. Participants undergoing general health check-ups or presenting with non-specific abdominal pain were included, provided they had both abdominal CT and USG imaging within a one-week timeframe. Exclusion criteria encompassed significant liver diseases, recent liver-related interventions, contraindications to CT imaging, and pregnancy. USG was performed using a Philips EPIQ 5G machine (Philips Ultrasound Inc., Washington, USA), grading FLD on a scale from Grade 0 (normal) to Grade III (severe). Subsequently, unenhanced CT scans were conducted using a Siemens 128-slice dual-source CT scanner (Siemens Healthineers, Munich, Germany), measuring liver attenuation values in HU. Statistical analyses included Spearman's rank correlation, analysis of variance (ANOVA), and receiver operating characteristic (ROC) curve analysis using IBM SPSS Statistics for Windows, Version 25 (Released 2017; IBM Corp., Armonk, New York, United States). Results The study population consisted of 60 (54.5%) males and 50 (45.5%) females with a mean age of 45.3 years and a mean BMI of 27.8 kg/m². USG grading revealed Grade 0 in 20 cases (18.2%), Grade I in 40 cases (36.4%), Grade II in 35 cases (31.8%), and Grade III in 15 cases (13.6%) of FLD. Mean CT HU values inversely correlated with USG grades, ranging from 65.2 ± 5.3 HU in Grade 0 to 35.6 ± 5.5 HU in Grade III. A significant negative correlation was observed between CT HU values and USG grading (Spearman's rho = -0.65, p < 0.001). CT demonstrated high diagnostic accuracy, with sensitivity and specificity of 80% and 85% for Grade I, 90% and 90% for Grade II, and 95% and 95% for Grade III FLD, respectively. The ROC analysis yielded an AUC of 0.85 (95% CI: 0.78-0.92) at an optimal cut-off value of 40 HU, achieving 90% sensitivity and 80% specificity in diagnosing moderate to severe FLD. Subgroup analyses indicated substantial agreement between CT and USG grading (Kappa = 0.72) and a significant association between higher BMI and increased FLD severity. Conclusion CT HU values exhibit a strong inverse correlation with USG grading of FLD and demonstrate high diagnostic accuracy, particularly in detecting moderate to severe steatosis. These findings support the use of CT as a reliable quantitative tool for FLD assessment, complementing the qualitative nature of USG. Integrating CT measurements into the diagnostic workflow can enhance the accuracy of FLD grading, especially in cases where USG limitations are present, thereby improving patient management and outcomes.