Adrenal lesions: variability in attenuation over time, between scanners, and between observers

Area of Science:

• Radiology
• Medical Imaging
• Oncology

Background:

• Computed tomography (CT) uses Hounsfield units (HU) and contrast wash-out to assess adrenal lesions.
• Variability in these measurements over time and between observers complicates adrenal lesion evaluation.

Purpose of the Study:

• To evaluate changes in adrenal lesion size, attenuation, and contrast wash-out over time using CT.
• To determine inter-observer variability and identify factors influencing measurement inconsistencies.

Main Methods:

• Prospective analysis of adrenal lesions in patients undergoing CT, excluding those with significant growth.
• Measurement of non-enhanced attenuation and contrast wash-out over a 2-year follow-up.
• Inter-observer analysis with five readers and a phantom study across eight CT scanners.

Main Results:

• Mean lesion attenuation decreased by 0.5 HU/year; 20-29% of lesions could be reclassified based on thresholds.
• High agreement (37/40 lesions) was observed between observers using established thresholds.
• Phantom studies revealed intra-scanner variability of 1-3 HU and inter-scanner variability up to 8 HU for water.

Conclusions:

• Strict reliance on specific attenuation threshold values for adrenal lesion characterization requires caution.
• Patient factors, equipment, scanning techniques, and observer influence impact measurement outcomes.