RSNA: Automated whole breast ultrasound feasible for cancer screening
CHICAGO—Automated whole breast ultrasound, which reduces dependence on operator skill, may be useful as a cancer screening tool, according to a Scientific poster presentation at the Radiological Society of North America (RSNA) annual meeting Sunday.
Alberto Martegani, MD, from Servizio di Diagnostica per Immagini, Ospedale Valduce, Como, Italy, and colleagues evaluated 32 women (with 35 lesions) from January to April 2009 with a CBUS 3000 (Helix Medical Systems) automated whole breast 3D device. Patients were placed in a prone position on an examination table and the ultrasound transducer was rotated 360 degrees in a radial fashion around the breast. Each rotation took approximately 60 seconds.
Reconstruction of the images into a 3D volume set was accomplished using a personal computer. The image layout included an axial, sagittal and coronal reformatted image plus the original image. The researchers evaluated one lesion per patient and compared the morphological aspect of the reconstructed images with the 2D acquired image.
The 35 lesions detected by free hand ultrasound were also detected by the 3D automated system. All lesions were evident on the “like original” images, while 10 and 12 lesions were identified on the sagittal and axial reformatted images, respectively. The coronal view identified 26 lesions.
The researchers concluded that while 2D “like original” views were more efficient in detecting lesions in images created by a 3D automated ultrasound system, combining the different views also gives an accurate topographic definition. “Automated whole breast ultrasound may have clinical utility as a cancer screening tool,” the authors said.
Alberto Martegani, MD, from Servizio di Diagnostica per Immagini, Ospedale Valduce, Como, Italy, and colleagues evaluated 32 women (with 35 lesions) from January to April 2009 with a CBUS 3000 (Helix Medical Systems) automated whole breast 3D device. Patients were placed in a prone position on an examination table and the ultrasound transducer was rotated 360 degrees in a radial fashion around the breast. Each rotation took approximately 60 seconds.
Reconstruction of the images into a 3D volume set was accomplished using a personal computer. The image layout included an axial, sagittal and coronal reformatted image plus the original image. The researchers evaluated one lesion per patient and compared the morphological aspect of the reconstructed images with the 2D acquired image.
The 35 lesions detected by free hand ultrasound were also detected by the 3D automated system. All lesions were evident on the “like original” images, while 10 and 12 lesions were identified on the sagittal and axial reformatted images, respectively. The coronal view identified 26 lesions.
The researchers concluded that while 2D “like original” views were more efficient in detecting lesions in images created by a 3D automated ultrasound system, combining the different views also gives an accurate topographic definition. “Automated whole breast ultrasound may have clinical utility as a cancer screening tool,” the authors said.