Effect of clinical history in focal liver lesion detection and classification on 4-detector row computed tomography and gadoxetic acid enhanced MR imaging in oncologic patients

To determine the influence of clinical data on reader diagnostic accuracy in focal liver lesion (FLL) detection and classification. Eighty-seven oncologic patients with FLLs underwent contrast-enhanced multidetector-row computed tomography (MDCT) and gadoxetic acid-enhanced magnetic resonance imaging (MRI). Two independent readers reviewed images for FLL detection and classification as benign or malignant without knowledge and after provision of clinical information. The sensitivity, specificity, diagnostic accuracy, and Az values were calculated using intraoperative ultrasound and pathologic findings as standard of reference. The awareness of clinical data significantly improved the accuracy (p = 0.02 for both readers) and Az values (p = 0.03 for reader 1 and p = 0.04 for reader 2) of MDCT for the detection of lesions with a diameter of 1 cm or less. When considering MRI, the provision of clinical data produced an improvement of accuracy and Az values for both readers, although differences were not significant.For MDCT lesion classification, the awareness of clinical data produced a decrease of accuracy (p = 0.03 for both readers) and Az values (p = 0.07 for reader 1 and p = 0.06 for reader 2) because of an associated increase in false-positive findings. When considering MRI, the provision of clinical data produced an improvement of accuracy and Az values for both readers, although differences were not significant. For detecting and classifying FLLs in oncologic patients, the knowledge of clinical data does not significantly change diagnostic accuracy and reader performance when using MRI, whereas, when considering MDCT, it improves the detection rate but produces an increase of false-positive findings for diagnosis of malignancy.