Focal liver lesion detection: concordance between volumetric US vs. conventional US
Purpose: To evaluate the accuracy of automatic volumetric acquisition in the detection of focal liver lesions in comparison to conventional US.
Materials and methods: 100 patients referred to our Institution underwent US of the superior abdomen. Operator first performed the study of the abdomen on conventional US, afterwhile volumetric acquisition of the liver was added. All examinations were performed with US Logiq9 (General Electric, USA). The order of acquisitions was inverted every new patient. Operator reported examinations on conventional US first. A second blinded operator read off line the volumetric acqusitions and reported imaging findings on the liver. Three categories were judged: 1=presence of focal liver lesions; 2=doubt of focal liver lesion; 3=absence of focal liver lesion. Concordance of volumetric acquisition to conventional US was evaluated with k test.
Results: 32 patients were reported to have benign focal liver lesions (19 cysts, 8 hemangiomas, 4 focal liver lesions in steatosis, 1 FNH) and 7 malignant (4 metastases, 3 HCC). In 61 patients no focal liver lesions were detected. From 3 to 7 acquisitions were needed to acquire the entire liver volume, 9s each, with a mean of 4,64 acquisitions/patient. All volumetric examinations were technically judged as good except of 5 with evident respiratory motion artefacts. A high concordance between volumetric acuisition and conventional US resulted (k=0.92)
Conclusions: Identification of focal liver lesion on volumetric acquisitions is a valid tool resulting high the concordance with conventional US.