Physica Medica: European Journal of Medical Physics - Volume 32 - Supplement 1 - February 2016

138 / 146

The final images were analyzed with TBSS (a FSL tool for Tract Based Spatial
Statistic), which can make an accurate comparison between values of frac-
tional anisotropy in white matter.
Results: The TBSS results present statistical positive differences between
the new preprocessing approach and the old one (p < 0.01, FWE cor-
rected) and none negative difference, showing a widespread improvement
of the contrast of FA with the adoption of the state of the art pipeline.
Conclusions: The importance of performing an accurate preprocessing of
the diffusion data is demonstrated, suggesting a careful revision of the old
pipeline used.
E.450
AIFMMULTICENTER INTERCOMPARISON OF MR SCANNERS FOR PROTON
SPECTROSCOPY – PRELIMINARY RESULTS
R. Sghedoni A. Coniglio G. Bell S. Buson A. Ciccaron M. Esposit
M. Giannelli L. Mazzoni L. Nocetti R. Tarducci I. Altabella
R. Anoja P. Berardi N. Bertolin C. Biagini M. Carnì P. Cesan
S. Cimolai S. Clemente E. Fabbr L. Fedel S. Filic F. Levrer
G. Meliadò N. Mordini S. Morzenti A. Moscat N. Oberhofe
N. Paruccini A. Ricci N. Romeo D. Scelf A. Toncell A. Torresin
M. Tosetti I. Zucca C. Gori 1 IRCCS-AO S. Maria Nuova, Reggio Emilia,
Italy; 2 San Giovanni Calibita – Ospedale Fatebenefratelli, Roma, Italy; 3 USL 8
di Arezzo, Arezzo, Italy; 4 AOU Careggi, Firenze, Italy; 5 AOU Meyer, Firenze, Italy;
6 Azienda Sanitaria Fiorentina, Firenze, Italy; 7 AOU Pisana, Pisa, Italy; 8 AOU
Senese, Siena, Italy; 9 AOU Policinico, Modena, Italy; 10 AOU Perugia, Perugia,
Italy; 11 Ospedale S. Raffaele, Milano Italy; 12 AO Pugliese-Ciaccio, Catanzaro,
Italy; 13 Policlinico S.Orsola-Malpighi, Bologna, Italy; 14 IRCCS- Istituto
Neurologico Besta, Milano, Italy; 15 Centro Oncologico Fiorentino, Sesto
Fiorentino, Italy; 16 Policlinico Umberto I, Roma, Italy; 17 AOU Città della Scienza
e della Salute, Torino, Italy; 18 ULSS 12 Veneziana, Mestre, Italy; 19 CROB, Potenza,
Italy; 20 AOU di Parma, Parma, Italy; 21 AOU S. Martino, Genova, Italy; 22 AOU
di Verona, Verona, Italy; 23 AO San Gerardo, Monza, Italy; 24AO Niguarda, Milano,
Italy; 25 Azienda Sanitaria dell’Alto Adige, Bolzano, Italy; 26 ASL di Viterbo,
Viterbo, Italy; 27 Azienda Sanitaria Provinciale di Messina, Messina, Italy;
28 IRCCS-Fondazione Stella Maris, Pisa, Italy
Introduction: The AIFM working group on MR intercomparison has pro-
posed a quality assurance protocol for magnetic resonance spectroscopy
(MRS). The aim of this study is to test the protocol on a significant number
of clinical MR scanners.
Materials and Methods: Ten centers and 12 MR scanners (9/3 1.5T/3.0T)
were enrolled in this study. A water phantom provided by the manufac-
turer of each MR scanner was used.
Three sets of unsuppressed water spectroscopy acquisitions were performed:
1) cubic voxel (size 20 mm) in the center of the phantom and in 2 posi-
tions at 40 mm from the center of the phantom in orthogonal directions;
2) in the center of the phantom setting 5 different echo times;
3) 5 acquisitions (voxel volume from 1 to 27 cm3).
The spectra were processed by jMRUI and the area under the water peak
(WPA) was calculated.
The following parameters were derived:
- from 1) the mean value of WPA, the standard deviation among the three
positions and the coefficient of variation (CV). The same parameters were
calculated for the peak width (PW);
- from 2) the relaxation time T2 of the phantom and the CV among sub-
sequent measurements;
- from 3) the linear correlation between both WPA and PW and the voxel
size.
Results: (a) CV of WPA as a function of spatial position was 13% on average
(<20% for 8 scanners over 12, > 50% in 3 scanners). The same was ob-
served for CV of PW (mean value 17% and just in 1 case > 40%); (b) phantom
T2 reproducibility was good for all the scanners (CV within 6% for all scan-
ners but 1); (c) WPA was highly linear as a function of voxel volume for
all the scanners, being the slope highly reproducible (within 10%) for all
the scanners but 1.
Conclusions: The proposed protocol provides a simple quality assurance
procedure in MRS. The derived parameters are generally reproducible. In
a new MRS intercomparison study already started, protocol acquisitions
will be performed by the participating centers using a standard phantom
containing metabolites.
E.451
QUANTITATIVE EVALUATION OF SUSCEPTIBILITY EFFECTS CAUSED BY
DENTAL MATERIALS IN HEAD MAGNETIC RESONANCE IMAGING
S. Strocchi E. Binaghi R. Novario M. Ghielmi F. Basilico a Fisica
Sanitaria, Ospedale di Circolo e Fondazione Macchi, Varese, Italy; b Dipartimento
di Scienze Teoriche e Applicate, Università della Insubria, Varese, Italy;
c Dipartimento di Biotecnologie e Scienze della Vita, Università della Insubria,
Varese, Italy; d Corso di Laurea in Tecniche di Radiologia Medica, per Immagini
e Radioterapia, Università della Insubria, Varese, Italy
Introduction: This work quantitatively evaluates the effects induced by sus-
ceptibility characteristics of materials commonly used in dental practice
on the quality of head MR images in a clinical 1.5 T device, in order to con-
tribute to the informed choice of dental material by dentists.
Materials and Methods: The proposed evaluation procedure measures
image artifacts induced by susceptibility in MR images by providing an index
consistent with the global degradation.
Susceptibility artifacts were evaluated in a near-clinical setup, using a
phantom with susceptibility and geometric characteristics similar to that
of a human head. We tested different dentist materials, called PAL Keramit,
Ti6Al4V-ELI, Keramit NP, ILOR F, Zirconia and used different clinical MR ac-
quisition sequences, such as “classical” SE and fast, gradient, and diffusion
sequences.
The evaluation is designed as a matching process between reference and
artifact affected images. The extent of the degradation is measured in terms
of similarity to the corresponding reference image. The matching process
involves a multimodal registration task and the use of an adequate simi-
larity index psychophysically validated, based on correlation coefficient.
Results and Conclusions: The proposed analyses are integrated within a
computer-supported procedure that interactively guides the users in the
phases of the evaluation method.
2-dimensional and 3-dimensional indexes were evaluated for each mate-
rial and each acquisition sequence. From these, we drew a ranking of the
materials, averaging the results obtained.
Zirconia and ILOR F appear to be the best choice from the susceptibility
artefacts point of view, followed, in order, by PAL Keramit, Ti6Al4V-ELI and
Keramit NP.
E.452
ARTEFACTS AND RELIABILITY IN 3D ATRIAL FIBROSIS SEGMENTATION OF
DELAYED-ENHANCEMENT CARDIAC MRI
A. Valentini D. Ravanelli E.C. Dal Pia M. Centonze G. Casagranda
M. Marini R. Bonmassar M. Del Greco a Medical Physics Department
– Azienda Provinciale per i Servizi Sanitari – APSS, Trento, Trento, Italy;
b Cardiology Department, S.Chiara Hospital – Azienda Provinciale per i Servizi
Sanitari – APSS, Trento, Trento, Italy; c Radiology Department – Azienda
Provinciale per i Servizi Sanitari – APSS, Trento, Trento, Italy; d Cardiology
Department, S.M. del Carmine Hospital, Rovereto – Azienda Provinciale per i
Servizi Sanitari – APSS, Trento, Rovereto, Italy
Introduction: For patients suffering atrial fibrillation and candidates for
catheter ablation, the clinical reference standard technique for left atrium
fibrosis assessment (LA-FA) is the use of electro-anatomical mapping system
(EAM). A promising alternative for LA-FA are delayed-enhancement MRI
(DE-MRI) images as a non-invasive and potentially more accurate method.
DE-MRI images are often of poor quality and present artefacts related to
non completed washout of contrast agent or non fibrotic related enhance-
ments. The robustness of algorithms against artefacts becomes also of
primary relevance. This study compares a semiautomatic threshold algo-
rithm with a fully automatic classifier.
Materials and Methods: 37 consecutive patients underwent DE-MRI and
were mapped with EAM. The atrial fibrosis percentage was calculated with
semiautomatic algorithm (threshold as 4 standard deviations above atrial
blood pool mean value) and with an automatic classifier (k-means clus-
e133 Abstracts/Physica Medica 32 (2016) e124–e134