This poster is published under an
open license. Please read the
disclaimer for further details.
Keywords:
Breast, Ultrasound, Diagnostic procedure, Image registration
Authors:
A. Morresi1, V. Girardi2, S. de Beni3, M. battaglia4, L. Forzoni5, A. martegani6; 1S. Fermo della Battaglia (CO)/IT, 2Brescia/IT, 3Genova/IT, 4Milano/IT, 5Firenze/IT, 6COMO /IT
DOI:
10.1594/ecr2016/C-1824
Results
The registration phase was performed selecting some easy to be recognized natural skin markers on the reference image and on the patient body ( such as nipple,
areola,
axilla ).
An average of 7 reference points was used in order to perform the registration between the real-time US and the reference image for 2D navigation.
The pointing and selection of the reference points was done directly with the US probe,
using the center of the array as pointer.
The same was done on the phantom using the marker placed on the surface.
It was always possible to use the 2D Navigation BodyMap Technology,
with a good correspondence between the probe representation (green circle) and the real probe position during real-time US.
In average the mismatch between the probe position and reference internal markers in in vivo,
ex vivo and in vitro were 13±9 mm.
Excessive compression during US scanning and Mammography acquisition created major distortions up to 22 mm.
The use of a proper amount of gel in order to ensure the correct coupling between the transducer and the body skin or phantom surface matching was useful in order to help to avoid excessive compressions caused by the transducer itself.
The registration time between US and any reference 2D image tested was 50 seconds in average.
Two “Digital” tests were performed regarding the check of the BodyMap capability to manage also simple (lateral compression) or complex (axial compression and lateral enlargement plus shift) of the examined body area or phantom.
The tests were performed on paper simulating one reference image and the related area to be scanned,
then comparing the BodyMap capability to mark the same targets on the original drawing and the compressed reference one.
The deformation was present in two dimensions (x,
y plane) the z axis defamation did not influence the result of the navigation since the second image modality did not have any spatial reference in this direction.
Complex distortions up to 2.5cm of compression axially and 1.7cm enlargement laterally were tested.