This poster is published under an
open license. Please read the
disclaimer for further details.
Keywords:
Extremities, Musculoskeletal system, Ultrasound, MR, Diagnostic procedure, Image registration
Authors:
M. Marinoni1, V. Saia1, T. Atzori1, S. de Beni2, S. D'Onofrio1, M. Olmi2, L. Forzoni1; 1Firenze/IT, 2Genoa/IT
DOI:
10.1594/ecr2015/C-1009
Methods and materials
Three (3) normal subjects were studied (ages 36,
39,
41; 1 female,
2 males).
All of the subjects could stand firmly for an average of 15 minutes during the acquisition of the weight bearing MRI images (the same time frame was required for the acquisition of supine MRI images).
A total of around 35 minutes,
including patient positioning,
was required for the full supine plus standing image acquisitions.
Twenty minutes were needed also for the fusion imaging session (with an extra 3 minutes required to upload the MRI study on the Ultrasound system second mode database for each subject).
MRI scans were performed using an Esaote GScan Brio system (0.25T,
permanent magnet technology),
see Fig.
2.
An appropriate 4-channel coil was used to assure a good and homogeneous magnetic field signal over the entire area of interest.
MRI femoral quadriceps images were acquired using a Spin Echo T1 sequence with a 5 mm slice thickness and a 0.5mm inter-slice distance.
Scout lines and proper Localizer image were previously acquired for correct slice positioning.
MRI scans were acquired in Axial and Sagittal plane to attain an optimal spatial image resolution of the two conventional Ultrasound scans.
Fusion imaging was performed using an Esaote MyLabTwice Ultrasound system equipped with a Linear array 10 MHz probe (LA533,
Esaote,
Italy) in conjunction with the Virtual Navigator fusion imaging tool (LA533 electromagnetic receiver support 639-042; CIVCO Medical Solutions,
Kalona,
Iowa,
USA).
The Virtual Navigator Ultrasound fusion system is equipped with an accurate algorithm which automatically selects the highest spatial resolution for the second mode image acquisition from the available group of data and according to the Ultrasound probe’s orientation.
The registration procedure was performed with 5 pinpoints that were placed on femoral quadriceps.
Registration was always successfully accomplished on first attempt (5 mm range).
A second fine tuning was required for half of the procedures mostly because of tissues’ bending/deformation due to MRI and Ultrasound different position acquisitions (due to MRI coil and leg possibly not being identically repositioned for MRI and Ultrasound scans).
Five (5) pinpoints (Fig.
3) were applied over the leg to ensure that all pinpoints were on different planes (one in the centre and four on the sides forming a square pattern).
Pinpoints were required as there are no anatomical landmarks available on the leg’s surface or within the leg’s structure (lack of vessel bifurcation or of specific muscular,
joint or bone signs).
Registration could theoretically be performed also taking into account leg’s internal planes/reference points.
A higher possibility of co-registration error,
at least on first attempt,
must be taken into account in this case.