Sources and impact of artifacts on clinical three-dimensional ultrasound imaging

Citation
Tr. Nelson et al., Sources and impact of artifacts on clinical three-dimensional ultrasound imaging, ULTRASOUN O, 16(4), 2000, pp. 374-383
Citations number
15
Categorie Soggetti
Reproductive Medicine
Journal title
ULTRASOUND IN OBSTETRICS & GYNECOLOGY
ISSN journal
09607692 → ACNP
Volume
16
Issue
4
Year of publication
2000
Pages
374 - 383
Database
ISI
SICI code
0960-7692(200009)16:4<374:SAIOAO>2.0.ZU;2-9
Abstract
The purpose of this paper is to investigate, identify and discuss artifacts and their sources arising in three-dimensional ultrasound (3D US) in clini cal practice in order to increase the awareness of clinicians and sonograph ers with respect to common 3D US artifacts and to use this increased awaren ess to avoid or reduce the occurrence of misdiagnosis in 3D US studies. Patient 3D US data were acquired using several different scanners and revie wed interactively on the scanner and graphics workstations. Artifacts were catalogued according to artifact origin. Two-dimensional ultrasound (2D US) artifacts were classified whether they were of a B-mode or color/power Dop pler origin and their presentation in the original scan planes and the resu lting volume re-sliced planes and rendered images was identified. Artifacts unique to 3D US were observed, noted and catalogued on the basis of whethe r they arose during acquisition, rendering or volume editing operations. Acoustic artifacts identified included drop-out, shadowing, etc. whose pres entation depended on the relationship between slice and imaging plane orien tation. Color/power Doppler artifacts were related to gain, aliasing, and f lash which could add apparent structure or confusion to the volume images. Rendered images also demonstrated artifacts due to shadowing and motion of adjacent structures, cardiac motion or pulsatility of the cardiac septum or vessel walls. Editing artifacts potentially removed important structures. Three-dimensional ultrasound is prone to the same types of artifacts encoun tered in 2D US imaging plus others unique to volume acquisition and visuali zation. The consequences of these diagnostically significant artifacts incl ude mimicking of abnormal development, masses, or missing structures thus r equiring careful study before reaching a diagnosis.