Background: A continuous pullback system is frequently used for acquisition of 3D-EUS images. However, free-hand systems have been used to obtain arbitrary volumes of data enabling flexible use of external ultrasound ( US ) probes. Aim: To test the performance in vitro and in vivo of a miniature magnetic sensor for endoluminal 3D-EUS. Methods: A miniature magnetic 6D position and orientation measurement system ( POM )( Mini-Bird TM ) was used to track different types of radial scanning endoluminal US probes. A precise six-degree of freedom digital position measurement table allowed a quantitative comparison of different probes and noisy influences from operating room environment.We used a non-ferric/non-magnetic endoscope ( MRI ), ( prototype MRI-endoscope, GF-MR30C, Olympus Co., Tokyo Japan ), with or without 12- ( UM2R ) and 20MHz ( UM3R ) miniature radial scanning US probes or an echoendoscope ( GFUM ). The performance is presented as mean difference between test values and control ( mm ± 2SD ), expressed as relative values, and/or correlation coefficients ( r ). 3D-EUS image reconstruction was evaluated on phantoms in vitro and finally in esophagus or stomach in anestesized pigs using frame-grabbed video images. Results: The receiver could be mounted on the endoscope tip or US scanheads. The 6D position an orientation measurement system allowed very good correlation (r=0,99), and was able to locate points with high accuracy (0,75 ± 2,60 mm), compared to absolute positions, over its specified operating range of 50 cm. The POM system was rather insensitive to errors resulting from eddy currents in conducting materials and interference from medical equipment near the receiver, typical 0,09 ± 0,20 mm, r=1,0. The relative performance when the POM was combined with different endoluminal probes was, UM2R (0,20 ± 0,42 mm, r=0,99), UM3R (0,20 ± 0,50 mm, r=0,99), MRI (0,85 ± 1,10 mm, r=0,99), GFUM (1.80 ± 1,88 mm, r=99. Software reconstruction based on free hand EUS scanning and POM registration resulted in accurate 3D-EUS images of the phantoms, gastrointestinal wall and adjacent structures that could be evaluated in a multiplane fashion. Conclusions: The performance of the POM system was similar if ferric or non-ferric probes were used for radial scanning EUS. 3D-EUS image reconstruction could be performed using free hand scanning in vitro and in animal organs. Combination of 3DEUS/POM registration may enable better description of true topographic relations and accurate volume measurements using 3D-EUS.
Read full abstract