Automatic Lesion Detection in Breast DCE-MRI

Abstract

Abstract. Dynamic Contrast Enhanced-Magnetic Resonance Imaging (DCE-MRI) has demonstrated in recent years a great potential in screening of high-risk women for breast cancer, in staging newly diagnosed patients and in assessing therapy effects. The aim of this work is to propose an automated system for suspicious lesion detection in DCE-MRI to support radiologists during patient image analysis. The proposed method is based on a Support Vector Machine trained with dynamic features, extracted, after a suitable pre-processing of the image, from an area preselected by using a pixel-based approach. The performance were evaluated by using a leave-one-patient-out approach and compared to manual segmentation made up by an experienced radiologist. Our results were also compared to other automatic segmentation methodologies: the proposed method maximises the area of correctly detected lesions while minimizing the number of false alarms (with an accuracy of 98.70%).

A Novel Model-Based Measure for Quality Evaluation of Image Registration Techniques in DCE-MRI

“…As suggested by Marrone et al [2] another perspective of the problem allows us to obtain a kind of approximated image registration by applying a median filter. The underlying hypothesis is that small transformations can be considered as noise superimposed on correctly aligned images.…”

“…For example, Marrone et al [2] proposed to apply a simple preprocessing based on median filter, assuming that small transformations can be considered as superimposed noise on baseline image (used as reference).…”

“…Dynamic Contrast Enhanced -Magnetic Resonance Imaging (DCE-MRI) is widely used as an important complementary diagnostic methodology for early detection of cancer [1] and for automatic lesion detection [2]. It demonstrated a great potential in screening of high-risk patients, in staging newly diagnosed cancers and in assessing therapy effects [3], thanks to its minimal invasiveness and to the possibility to visualize 3D high resolution dynamic (functional) information, not available with conventional RX imaging [4]- [6].…”

“…Its suitability in evaluating image registration techniques has been indirectly demonstrated by considering the results obtained, after each image registration technique, by a previously proposed automatic detection system for neoplastic lesions (BLADeSBreast Lesion Automatic Detection System [2], [14]). Results showed that the ranking obtained by means of the proposed quality assessment measure is in agreement with the ranking of the CAD segmentation systems, proving that our measure can correctly evaluate image registration techniques in the DCE-MRI context.…”

“…One of the main issues in developing a CAD system is the detection of suspicious Region Of Interests (ROIs) as sensitively as possible, while simultaneously minimising the number of false positives [2], [8], [9]. This task is made harder by the dynamical characteristics of the DCE-MRI examinations of soft tissues: patient movements (due to inspiration or voluntary and involuntary movements) may affect the voxelby-voxel dynamical analysis (Figure 1 -left).…”