LSN Score - Liver Surface Nodularity Score

How It Works

• Imaging software detects the liver surface contour on CT
• The LSN score = average distance between the detected liver surface and a smoothed polynomial "best-fit" line
• 10 contiguous CT slices are evaluated (total distance: 80-100 cm)
• Final score = average of all slice scores, expressed in tenths of a millimeter
• Average processing time: <2 minutes

Score Range & Interpretation

• In NAFLD studies: Mean LSN - Normal controls: 1.30 ± 0.09, Simple steatosis: 1.54 ± 0.21, NASH: 1.59 ± 0.23
• Mean LSN by fibrosis grade: F0 = 1.30, F1 = 1.45, F2&F3 = 1.67

Diagnostic Accuracy for Hepatic Fibrosis Staging

• In NAFLD specifically: Cut-off LSN >1.48 distinguishes F1 from F2/F3 with sensitivity 83.3%, specificity 72.7%, AUROC 0.788
• Accuracy further improves when combined with FIB-4 index (age + AST + ALT + platelet count)

Reliability

• Inter-observer agreement: ICC 0.94 (CV 9%) - very high
• Test-retest reliability: ICC 0.82 (CV 12%)
• Software-based LSN has higher inter-observer agreement than subjective visual scoring (ICC 0.84 vs 0.61)
• Technical failure rate: approximately 0-3.9%

Uses

1. Non-invasive fibrosis staging in NAFLD, NASH, HCV, and other chronic liver diseases
2. Cirrhosis sub-staging - predicts hepatic decompensation and death
3. Monitoring disease progression over time
4. Reduces need for liver biopsy in many patients

Context Among Other NAFLD Scoring Tools

• NFS (NAFLD Fibrosis Score) - uses age, BMI, hyperglycemia, AST/ALT, platelets, albumin
• FIB-4 - age, AST, ALT, platelets
• BARD score - BMI, AST/ALT ratio, diabetes
• LSM (Liver Stiffness Measurement) via FibroScan/transient elastography (in kPa)

Liver Surface Nodularity (LSN)

Pathophysiology

Detection Methods

1. Ultrasound (Qualitative)

• Right lobe atrophy with compensatory left lobe/caudate hypertrophy
• Parenchymal inhomogeneity
• Signs of portal hypertension (splenomegaly, varices, ascites)

2. CT - Quantitative LSN Score (Primary Method)

• Imaging software automatically detects the liver surface outline on axial CT slices
• LSN score = average distance between the detected liver surface and a smoothed polynomial "best-fit" line
• 10 contiguous CT slices are evaluated (total coverage: 80-100 cm)
• Final score = average of all slice scores, expressed in tenths of a millimeter
• Processing time: <2 minutes
• Works on non-contrast and contrast-enhanced CT

Diagnostic Accuracy for Fibrosis Staging (CT-based)

• In NAFLD specifically: cut-off LSN >1.48 differentiates F1 from F2/F3 with sensitivity 83%, specificity 73%, AUROC 0.788
• Accuracy further improves when combined with FIB-4 index (age + AST + ALT + platelet count)

Clinical Applications

1. Non-invasive fibrosis staging in chronic liver disease (HCV, NAFLD/NASH, alcohol-related liver disease)
2. Substaging of cirrhosis - the LSN score is a strong predictor of hepatic decompensation (variceal bleeding, encephalopathy, ascites) and death (Smith et al., Radiology 2017)
3. Portal hypertension detection - LSN correlates with hepatic venous pressure gradient (Sartoris et al., Radiology 2018)
4. Reducing liver biopsies - high negative predictive value for ruling out advanced fibrosis
5. Retrospective staging from existing CT images (no additional scan needed)

Reliability

Limitations and Contraindications

MRI-Based LSN

Context Among Non-Invasive Fibrosis Tools