A Gradient-Based Approach for Breast DCE-MRI Analysis

Abstract

Breast cancer is the main cause of female malignancy worldwide. Effective early detection by imaging studies remains critical to decrease mortality rates, particularly in women at high risk for developing breast cancer. Breast Magnetic Resonance Imaging (MRI) is a common diagnostic tool in the management of breast diseases, especially for high-risk women. However, during this examination, both normal and abnormal breast tissues enhance after contrast material administration. Specifically, the normal breast tissue enhancement is known as background parenchymal enhancement: it may represent breast activity and depends on several factors, varying in degree and distribution in different patients as well as in the same patient over time. While a light degree of normal breast tissue enhancement generally causes no interpretative difficulties, a higher degree may cause difficulty to detect and classify breast lesions at Magnetic Resonance Imaging even for experienced radiologists. In this work, we intend to investigate the exploitation of some statistical measurements to automatically characterize the enhancement trend of the whole breast area in both normal and abnormal tissues independently from the presence of a background parenchymal enhancement thus to provide a diagnostic support tool for radiologists in the MRI analysis.

A Clifford Analytic Signal-Based Breast Lesion Segmentation Method for 4D Spatial-Temporal DCE-MRI Sequences

“…For the QIN DCE dataset, each study contained more than 20 time points of DCE-MRI sequences, and all were used to compute the CTI. Since the subtraction image is used in conventional methods [16], [19], [21]- [23], the post-contrast sequence at peak enhancement time point and the pre-contrast sequence were used to obtain the best subtraction image in our experiment. Figure 3 shows two examples of a CTI and DCE-MRI subtraction image on one slice from the QIN DCE dataset.…”

“…For a fair comparison with existing subtraction imagebased methods [16], [19], [21]- [23], the breast ROI and subtraction image were used to train another lesion segmentation FCN network; we call this the DCE-MRI subtraction image-based (DS) method in this paper. In addition, the DeepMedic (DM) method [31] was adopted as a whole chain method for comparison.…”

“…However, one subtraction image cannot eliminate enhanced vessels and fibroglandular tissue because of BPE. The problem is somewhat alleviated based on the temporal gradient information by manually selecting one best time point of the DCE-MRI sequence [23], but manual selection is time-consuming.…”

“…Approaches with manual pre-selection of lesion ROI and type (mass and NME) are generally inconvenient in real clinical applications. Moreover, approaches that do not use multiple or sufficient time points of DCE-MRI sequences may overlook important temporal information for lesion segmentation [16], [19], [21]- [23]. An end-to-end framework that can fully utilize the temporal information from multiple DCE-MRI data to segment both mass-type and NME lesions is still lacking.…”

Radiomics Analysis on Contrast-Enhanced Spectral Mammography Images for Breast Cancer Diagnosis: A Pilot Study

“…Among these new techniques, Contrast-Enhanced Spectral Mammography (CESM) [ 31 , 32 , 33 ] combines the principles of mammography with the injection of an intravenous iodinated Contrast Medium (CM), which allows, as in MR images, a contrastographic evaluation of the breast: this highlights the areas that capture the contrast medium, the typical expression of neo-angiogenesis neoplasm. As in MR, CESM images may present enhancement of normal breast parenchyma after CM injection, known as Background Parenchymal Enhancement (BPE) [ 15 , 34 ]. On the contrary, CESM is less influenced by hormonal status than MR [ 35 ], and this could provide important additional information on the detection of lesions in patients with a high BPE in degree where distinguishing a lesion from the non-enhanced background is objectively difficult.…”