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

92 / 146

Purpose: The Ligurian Medical Physicist working group on Mammo-
graphic Screening analyzed the results of two quality controls, homogeneous
in all centers, in order to provide to the regional network HTA a compre-
hensive indicator as clinical parameter within the obsolescence evaluation
of a system.
Methods and Materials: The average glandular dose (AGD) and the thresh-
old contrast visibility were chosen as quality indicators. The results of these
tests are represented by a curve that must remain below the curve of ac-
ceptable level and, if possible, also below the curve of the achievable
level.
In order to obtain two indicators, dose and contrast one, the average dis-
tances between the measured and acceptable values has been calculated
for both the physical quantities and then normalized to acceptable values.
Then dose (DI) and contrast indicators (CI) DI and CI have been linked into
an a-dimensional index.
Results: All the mammography systems deliver an AGD below the accept-
able levels. Depending on DI definition, the higher the value the lower is
the dose compared to the acceptable value. It was noted that the CR devices
deliver an AGD higher than DR ones.
CI showed that CR equipment has a contrast-detail curve worse than the
DR ones. For some mammography (due to some points out of tolerance)
the CI was even below the acceptable level.
Conclusions: These two tests are not sufficient to comprehensively eval-
uate the equipment, but are indicative of the dose and image quality and
they were uniformly implementable in all the centers.
The overall assessment parameter that was provided to the Ligurian network
HTA will be part of a set of indexes, such as technical, functional and eco-
nomical obsolescence, which will together give advice relating to any
mammography equipment obsolescence.
This benchmark could allow the Ligurian network HTA to suggest in the
future the purchase of equipment with CR technology and to assess which
device prioritizes for upgrades or replacements.
B.294
MULTI HOSPITAL EXPERIENCES OF RADIATION DOSE INDEX
MONITORING
S. Pini S. Mazzocchi A. Bruschi A. Ghirelli E. Rigacci G. Zatell
a Azienda Sanitaria Di Firenze, Firenze, Italy; b Universita’ Degli Studi Di Firenze,
Firenze, Italy
Introduction: Radiation Dose Index Monitoring (RDIM) software is an im-
portant tool to optimize radiological procedures, however its implementation
need a thorough acceptance test.
Methods and Materials: In our multi hospital reality (ASF) we begun in
2012 to monitor dose index data from 7 CTs and 2 angiographic units using
CareAnalytics (CA) a non-medical, free-of-charge tool by Siemens. CA pro-
cesses DICOM Radiation Dose Structured Reports (RDSR) stored in PACS.
Moreover in context of an AIFM project, Radiation Dose Monitor (RDM) by
Medsquare has been tested since Gen 2015. RDM is a web based software
solution that collects, controls and analyzes radiation doses delivered to
patients during medical imaging examinations, it supports either modal-
ity integration, or RIS and Pacs integrations and can retrieve dose and
relevant data from RDSR, directly from header DICOM or via MPPS mes-
sages. In ASF RDM server has been connected to the single modalities in
the RIS PACS network: one mammographic installation, 2 CTs and 1
angiographic unit. The aim of this work is to test the consistency of data
processed by either two RDIM software.
Results: For angiographic procedure we have compared the maximum in-
cident air kerma at the reference point at different times and at different
C arm positions with the dose values obtained by software analysis. For
mammography, data from header DICOM (kV, HVL, breast thickness, AGD)
have been positively compared with RDM data. For CT we have compared
values obtained from console, CA and RDM.
Conclusion: In our experiences both the software under test seems to be
useful tools to monitor radiation dose index. RDM allows wider applica-
tions and custom made options.
B.295
THREE YEARS OF EXPERIENCE WITH DOSE REDUCTION BY MEANS OF
ITERATIVE RECONSTRUCTION ALGORITHM AVAILABLE ON CT SIEMENS
SOMATOM DEFINITION FLASH
M. Poli *, C. Bracco, S. Bresciani, A. Maggio, A. Di Dia, A. Miranti,
M. Stasi. Medical Physics, Candiolo Cancer Institute, FPO-IRCCS, Candiolo (TO),
Italy
Introduction: Patients in a radiological unit undergo multiple CT scans, re-
sulting in high cumulative radiation exposures. Three years of experience
with dose reduction by means of CT Somatom Definition Flash is here pre-
sented. In particular, the implementation of SAFIRE low-dose CT protocols
and the results in terms of dose (D) are shown. New protocols were first
proposed by the physicist after a phantom study, then the selected pro-
tocols were clinically implemented only after a clinical validation with which
the physician confirmed no loose of diagnostic quality image.
Material and Methods: Two acquisition protocols were considered: the
thorax-abdomen (TA) and the head (H). Both protocols were optimized and
the new protocols were clinically implemented in the 2014. The recon-
struction kernel was changed from the iterative I30f strength 2 to the I30f
strength 3 for the TA protocol and from the filtered back projection H31s
to the iterative J40s strength 3 for the H protocol. Retrospectively ana-
lyzed were CTDIvol of 100 patients for each year (2013–15) and for each
protocol in order to estimate the dose reduction gained since the intro-
duction of these new protocols.
Results: The average CTDIvol relative to the pre-optimization protocol was
(13.7 ± 1.8) mGy for the TA and 60.8 mGy for the H protocol. The TA pro-
tocol optimization led to a dose reduction between 2013 and 2014 of the
30%: the average CTDIvol relative to 2014 and 2015 was (9.9 ± 1.5) mGy
and (9.7 ± 0.6) mGy, respectively. Similarly, we obtained a dose reduction
of 10% for the H protocol, equivalent to an average CTDIvol of 54.7 mGy
in the 2015.
Conclusions: Protocols optimization by the use of iterative reconstruc-
tion methods, such as SAFIRE, can lead to an important dose reduction.
B.296
ITERATIVE METHOD FOR CT SYSTEM: DOSE REDUCTION AND
QUANTITATIVE ANALYSIS OF IMAGE QUALITY IMPROVEMENT
M. Porzio *, F. Coloberti, E. Zucchi. ASL 1 Imperiese, Sanremo, Italy
Introduction: This work analyzes a commercial iterative method (Adap-
tive Statistical Iterative Reconstruction-ASiR – GE Medical Systems) for which
noise reduction technique may allow for dose reduction without loss of
image noise levels or be applied to improve images with unacceptable level
of noise.
Materials and Methods: Three clinical protocols are used on Optima
CT540 system: axial Routine Head, Helical Head, Abdomen Pelvis. DCT10
and CTSD16 RTI probes in PMMA phantoms for head and body CTDI
measurements are used for each clinical protocol at the available dose
reduction levels (0, 20, 30, 40%). The measured CTDI are compared with
the expected reduction. A Cathpan 600 phantom is scanned at the
above levels and the analysis is performed with AutoQA Lite™-IrisQA
software. Furthermore, ASiR is tested on image data obtained at 0% dose
reduction level, applying retrospectively all the available steps of noise
reduction.
Results: The measured CTDI confirm the nominal dose reduction levels and
only at the upper dose reduction, noise level shows an increase of 7% in
the axial Routine Head protocol, while for all clinical protocols uniformi-
ty and low contrast detectability (LCD) are unchanged and MTF50 improves
up to 20%. Referring to quality of reconstructed images, depending on the
steps of noise reduction, all parameters are improved: MTF50 grows up to
23% in Abdomen Pelvis protocol, noise is reduced to 33% in Helical Head
protocol, LCD increases of 43% in Abdomen Pelvis protocol, while unifor-
mity is always unchanged.
Conclusion: CT extensive use in medical procedures contributes about 50%
of the population’s medical exposure. The raising radiation dose to the
patient population from CT has created a major concern in the medical
imaging community in an effort to optimize procedures and to reduce ra-
diation risk.
e87 Abstracts/Physica Medica 32 (2016) e71–e96