The digital revolution has made three-dimensional (3D) ultrasonography a natural extension of 2D ultrasound scanning. This lecture reviews some of the principles and methods in 3D ultrasonography of the upper abdomen. At present, the process of making 3D images based on ultrasonography is commonly divided into 5 steps: Data acquisition, data digitization, data storage, data processing, and data display. Principally, data acquisition by 3D ultrasonography can be performed in 3 different ways; either by using a 2D probe attached to a motor which moves the probe in a computer-defined way or by a spatial localizing system connected to a 2D probe or by genuine, electronic 3D probes with the possibility of direct volume acquisition in real time. True volumetric 2D array transducers instantly generate a volume of ultrasound data, enabling dynamic 3D ultrasonography. A critical step in processing of 3D data is segmentation, which is the procedure where the object of interest is separated from the surrounding structures. Three fundamental approaches to segmentation have been utilised in 3D ultrasonography: Extraction by visual inspection and manual outlining of contours; semi-automatic separation using visualisation algorithms aided by operator interaction, and fully automatic computer segmentation. The final step in formation of 3D ultrasound images is to display the data so that the inherent voxel information is communicated accurately. Commonly, simple rotation of the object on the computer monitor provides some 3D effects. To further enhance the 3D outcome of the images, stereoglasses can be applied. Data processing and display requires specialized computer software to handle the ultrasound images. Ultrasound data contains a significant amount of noise and speckle and may exhibit boundary regions several pixels wide. It is important to keep in mind that the quality of the final 3D data display strongly depends on the resolution of the raw data. Transducer frequency and lateral resolution, frame rate of the scanner, accuracy of 3D probe, speed of scanning, methods of filtering and segmentation, are all factors that influences the final 3D image and subsequently accuracy in volume measurements.
Read full abstract