Effect of slice thickness on liver lesion detection and characterisation by multidetector CT

The purpose of our study was to compare the effectiveness of 3.2 mm, 5 mm and 7.5 mm slice thicknesses in the detection and characterisation of liver lesions found on CT in patients with known or suspected malignant disease. 110 patients underwent portal phase imaging using four-slice MDCT. Two blinded observers independently read hard copy images at each slice thickness. The size and location of each lesion detected was recorded by each observer on a diagram of liver segmental anatomy. Each lesion was characterised as benign, malignant or indeterminate in nature. A diagnostic confidence score was allocated for each lesion on a scale of 1-4. The pathology or behaviour of lesions was assessed using surgery with intra-operative ultrasound (IOUS) and histology, or interval imaging with MRI, CT, or sonography. 294 lesions were detected, 64 (22%) of which were malignant. Both observers detected significantly more lesions on the 3.2 mm versus 7.5 mm slice thickness (p < 0.0001). Both observers detected more malignant lesions on 3.2 mm and 5 mm slice thicknesses versus 7.5 mm. As slice thickness decreased there was a significant increase in the sensitivity of malignant lesion detection for observer 1 (p < 0.001) and borderline significance for observer 2 (p = 0.07). As slice thickness decreased the proportion of lesions characterised as indeterminate by both observers fell. With thinner slices, both detection and characterisation of liver lesions were improved. A slice thickness no greater than 5 mm should be used to maximise both detection and correct characterisation of liver lesions.