[Tridimensional ultrasonography. First clinical experience with dedicated devices and review of the literature].

PURPOSE

We report our preliminary clinical experience with three-dimensional ultrasound (3D US) in abdominal and small parts imaging, comparing the yield of 3D versus 2D US and through a literature review.

MATERIAL AND METHODS

We used a Tomtec Echo-Scan 3.1 connected to a Philips P 700 unit with a 3.5 MHz convex probe and to a Toshiba SSA-340 A (equipped with power Doppler) with a 3.5 MHz convex and a 7.5 MHz linear probes. The system consists of: a) a workstation (166 MHz Intel Pentium, 128 Mbytes RAM, 520 Mbytes hard disk, 1.3 Gbyte Magneto-Optical drive); b) a spatial location system (3D Freehand Scanning) whose sensor, attached to the probe, provides spatial coordinates for each US scan in an electromagnetic field created by a transmitter; the software can thus correctly stack 2D US images to make 3D reconstructions of anatomical structures. The technical steps are: 1) setting; 2) image acquisition; 3) image processing and 3D rendering using surface or volume modes; 4) image archiving. 2D US was performed on 50 subjects, namely 20 volunteers and 30 patients with different pathologic conditions and 3D reconstructions were obtained from the best US images. We evaluated which anatomical structures and pathologic conditions are best suited for 3D rendering.

RESULTS

The best 3D images were obtained from anatomical structures and pathologic conditions with a liquid content (i.e., bladder and gallbladder; cysts), or those adjacent to them (i.e., uterus and prostate). Major limitations were encountered in the assessment of the parenchyma of liver, kidneys, pancreas, thyroid, testis and breast, due to intrinsic texture low contrast, while intraparenchymal liquid structures (i.e., vessels, urinary cavities) and structures surrounded by liquid (i.e., hydrocele, ascites) were better demonstrated.

DISCUSSION

The system permits accurate spatial location, and therefore stacking, of each US scan; this provides good-quality 3D images with fewer artifacts. The system can be connected to any existing US unit and to many kinds of probes. Incorrect processing or rendering may worsen 3D image quality and thus anatomical reconstructions; other drawbacks may come from difficult stacking of reconstructed images or limited field of view. Our personal experience and the review of 3D US literature indicate that the system may be used for the following clinical applications: anatomical assessment of lesions for minimally invasive treatment; targeting areas of interest and adjacent structures during radiotherapy; lesion volume studies during therapy; 3D vascular mapping with power Doppler; 3D reconstructions by intraluminal approach; real-time 3D scanning for US guidance during minimally invasive procedures.

CONCLUSIONS

Our preliminary experience suggests that technological progress will soon lead to a widespread use of 3D US and its applications.