ABSTRACT
PURPOSE: To compare video sequence and conventional freeze image documentation of breast ultrasound findings with respect to lesion assessment and the diagnostic power of established ultrasound characteristics. MATERIALS AND METHODS: Digitally stored freeze images and corresponding video sequences of 50 breast lesions were randomly arranged and interpreted by eight investigators with breast ultrasound training and experience. Established ultrasound criteria were documented on a standardized classification form for every lesion. The investigators were blinded to the clinical and radiological findings, patient characteristics including age, and lesion histology. Statistical analysis compared both groups and correlated the results with the lesion histology. A receiver operating characteristics (ROC) analysis was performed to evaluate the diagnostic performance of ultrasound criteria in lesions documented by video sequences compared to freeze images. RESULTS: Breast lesions were assessed almost identically in video sequences and freeze images. Only the features echogenic halo, orientation, and margin varied among both groups. The dynamic features compressibility and mobility were highly statistically significant correlated with the lesion histology in the video sequence interpretation (p < 0.0001). ROC analysis revealed almost identical diagnostic accuracy in both groups (area under the curve 0.719 for video sequences and 0.762 for freeze images). CONCLUSION: Video sequences are an appropriate tool to document lesions in breast ultrasound. In contrast to our hypothesis, however, this tool did not improve the diagnostic power of established ultrasound characteristics compared to freeze image documentation.
