Ultrasound is NOT Superior to CT for Assessing Fatty Liver
Neither ultrasound nor CT should be considered first-line modalities for accurate assessment of fatty liver—MRI is superior to both—but when comparing these two specifically, CT outperforms ultrasound for quantification while ultrasound has limited diagnostic accuracy.
Key Evidence Against Ultrasound
The most critical finding from systematic review evidence is that ultrasound has a positive predictive value of only 47-62% for fatty liver in children, failing to meet standard clinical diagnostic thresholds 1. This poor performance stems from ultrasound's fundamental limitation: it does not directly measure fat but relies on subjective, non-quantitative interpretation of echogenicity.
Specific Ultrasound Limitations:
- Cannot grade steatosis accurately: Children with "mild" steatosis by ultrasound had moderate steatosis on histology in ~50% of cases 1
- High false positive rate: When compared to MR spectroscopy, 72% of patients with mild steatosis by ultrasound (score of 1) actually had no fatty liver 1
- Poor correlation with quantitative measures: Only 48% of ultrasound-positive cases were confirmed by MRI 1
- Cannot exclude fatty liver reliably: Most negative ultrasounds in one study were falsely negative 1
A meta-analysis found ultrasound sensitivity of 84.8% and specificity of 93.6% for moderate-to-severe fatty liver only 2, meaning it misses mild disease entirely.
CT Performance
CT should not be used for primary assessment of hepatic steatosis due to ionizing radiation exposure 1. However, when CT is performed for other indications:
- Quantitative capability: CT provides objective Hounsfield Unit (HU) measurements that inversely correlate with fat content 3, 4
- Diagnostic thresholds: Portal venous phase liver attenuation ≤80 HU has 85.6% sensitivity and 89.6% specificity for moderate steatosis 4
- Post-contrast CT alone (liver HU) outperforms liver-spleen difference (AUROC 0.943 vs 0.814) 4
- Unenhanced CT threshold: Liver attenuation <40 HU corresponds to >15% fat fraction 4
However, CT has significant limitations:
- Results confounded by cirrhosis and iron overload 5, 6
- Iron overload dramatically reduces CT sensitivity: Only 20% sensitivity with dual-energy CT in iron-overloaded patients 5
- Radiation exposure precludes use for screening or monitoring 1
Clinical Algorithm
For fatty liver assessment:
First-line: MRI with proton density fat fraction (PDFF) techniques—most accurate and quantitative 1
If MRI unavailable or contraindicated:
- Use ultrasound only for initial screening in low-risk populations
- Recognize that negative ultrasound does NOT exclude fatty liver
- Any positive ultrasound requires confirmation with MRI or biopsy
If CT already performed for other indications:
Never order CT primarily for fatty liver assessment due to radiation 1
Critical Pitfalls
- Do not use ultrasound to grade steatosis severity—it lacks accuracy for this purpose 1
- Do not trust liver-spleen attenuation differences on CT—direct liver HU is superior 4
- Do not use either modality in patients with suspected iron overload—both perform poorly 5
- Ultrasound's widespread use is based on convenience and cost, not accuracy 1, 2
Bottom Line
Ultrasound is inferior to CT for quantifying fatty liver, but CT should not be used as a primary diagnostic tool due to radiation. The 2014 systematic review definitively showed ultrasound "did not meet the standard clinical threshold required to be used as a diagnostic test" 1. When comparing the two directly, CT provides objective quantification while ultrasound remains subjective and inaccurate, particularly for mild-moderate disease. MRI remains the preferred non-invasive modality when accurate assessment is clinically necessary 1, 7.