This poster is published under an
open license. Please read the
disclaimer for further details.
Keywords:
Breast, Mammography, Diagnostic procedure, Cancer
Authors:
P. MARTÍNEZ MIRAVETE1, P. Garcia Barquin2, M. Millor Muruzábal2, J. Etxano2, A. Elizalde2, L. J. Pina Insausti2, L. R. Zalazar2; 1Zaragoza/ES, 2Pamplona/ES
DOI:
10.1594/ecr2015/B-0923
Methods and materials
Patients selection
From September 2013 to September 2014 we began to use the synthesized-DM and the combo mode.
A total number of 8201 COMBO studies were performed at our institution in Zaragoza (Spain).
The informed consent was offered to all patients.
All the studies were performed using a Selenia Dimensions unit (Hologic,
USA).
Study design
We retrospectively selected an enriched sample with 210 patients.
All patients underwent both techniques 2D and synthesized DM.
The reading modes included standard craniocaudal (CC) and mediolateral-oblique (MLO) views of each breast.
In our series,
57 patients had histologically proven malignant lesions and 153 patients had benign lesions (54 patients with histologic confirmation and 99 benign lesions with 1 year follow up without changes).
One expert breast radiologist,
blinded to the clinical information and histological diagnosis,
reviewed all the studies.
Firstly 2D-DM and four weeks later Synthesized-DM studies were evaluated.
This study was designed as a patient-based study,
so additional suspicious lesions were excluded.
For each case,
the reader evaluated the 2D-DM and synthesized-DM and classified them according to the BI-RADS categories,
as negative (BI-RADS 1 to 2) or positive (BI-RADS 3 to 5).
Statistical analysis
The differences in Sensitivity (SE) and Specificity (SP) between 2D-MD and Synthesized-DM were calculated using PEPI software.
The ROC curves were calculated and compared using a z test and SPSS (20.0).