Richter K, Prihoda H, Heywang-Köbrunner S H, Hamm B
Department of Radiology, Charité Medical School, Humboldt University of Berlin, Germany.
Invest Radiol. 1997 Jan;32(1):19-28. doi: 10.1097/00004424-199701000-00004.
The authors tested a new automated system for clinical amplitude/velocity reconstructive imaging (CARI) breast sonography and compared the results with those of a manual laboratory model of this imaging modality, conventional sonography, and mammography.
Fifty-one patients with breast lesions were examined by mammography, conventional breast sonography, a laboratory version of a mammography-like breast compression device for breast sonography, and an automated system that allows sonography while in compression and reconstructs an image of the whole breast. The automated device allows reconstruction of one image per breast from an automatically acquired sonographic data set and simultaneous performance of mammography. The results of mammography and all modes of sonography were compared with histopathology in 36 surgically treated lesions of which 20 were benign and 16 malignant. In addition, 17 lesions confirmed to be benign by different imaging modalities were examined with the three different sonographic devices.
All 16 histopathologically proven malignant lesions were identified correctly by compression sonography (laboratory and automated device) compared with 14 by conventional sonography. Of the 20 benign lesions, 16 were interpreted correctly by conventional sonography, 19 with the laboratory version, and 18 with the automated device, whereas mammography identified 12 benign lesions correctly and missed 1 malignant lesion. Only 16 of 17 lesions diagnosed to be benign by outpatient imaging could be examined with all three modes of study sonography and were found to be benign with all three modes in all cases.
Compared with the laboratory version, the automated device facilitates acquisition and evaluation of compression breast sonography and is similar in lesion detection and differentiation. It should be tested in a larger number of patients.
作者测试了一种用于临床振幅/速度重建成像（CARI）乳腺超声检查的新型自动化系统，并将结果与该成像方式的手动实验室模型、传统超声检查和乳腺钼靶检查的结果进行比较。
对51例乳腺病变患者进行了乳腺钼靶检查、传统乳腺超声检查、用于乳腺超声检查的类似乳腺钼靶的乳腺压迫装置的实验室版本检查，以及一种允许在压迫状态下进行超声检查并重建整个乳腺图像的自动化系统检查。该自动化装置可从自动采集的超声数据集中为每个乳房重建一幅图像，并同时进行乳腺钼靶检查。对36例手术治疗病变（其中20例为良性，16例为恶性）的乳腺钼靶检查和所有超声检查模式的结果与组织病理学结果进行比较。此外，对通过不同成像方式确诊为良性的17个病变，用三种不同的超声检查设备进行检查。
与传统超声检查发现14个相比，所有16个经组织病理学证实的恶性病变通过压迫超声检查（实验室和自动化装置）均被正确识别。在20个良性病变中，传统超声检查正确解读了16个，实验室版本正确解读了19个，自动化装置正确解读了18个，而乳腺钼靶检查正确识别了12个良性病变，漏诊了1个恶性病变。门诊成像诊断为良性的17个病变中，只有16个能用所有三种研究性超声检查模式进行检查，且在所有病例中所有三种模式均显示为良性。
与实验室版本相比，该自动化装置便于进行压迫乳腺超声检查的采集和评估，在病变检测和鉴别方面相似。应在更多患者中进行测试。
