Breast cancer represents the most common oncological disease for women and a multimodality approach, with a clinical exam and an advanced set of BI examinations, is mandatory as current medical standard.
MRI and Contrast-enhanced MRI (CE-MRI) play a central role in BI, due to their very high sensitivity. In addition, technology evolution over the recent years has made breast US an essential component of the BI evaluation. In current praxis, when lesions detected on MRI are not visible/equivocal on mammogram and/or are not identified by previous US, a second-look US is commonly prescribed, to provide additional lesions information or lesion precise localization. Obviously said, second-look US examination must be correlated with previous MRI findings.
Breast tissues are soft and easily deformable, the organ is movable and breast size varies substantially among women; in addition, Breast MRI is routinely performed with patients in prone position, while US is performed with the patient in supine position. As a result of varying patient’s position during clinical examinations, breast location, its size and the localization of potential internal lesions typically undergo significant variations, with substantial spatial displacement and misalignment.
A new Fusion Imaging (FI) algorithm has been developed (BreastNav®, MedCom GmbH) and embedded on MyLab 9 US system (Esaote SpA, Italy), helping to overcome such difficulties and to provide all the diagnostic and clinical advantages of BI MRI-US Multimodal Fusion.
BreastNav® allows to correlate prone MRI to supine US and to localize on the real time US, with patient supine, the spatial position of a reference anatomical target, related to a lesion under investigation, previously identified on prone-position MRI.
The algorithm needs minimal user input and interaction and performs automatically a mathematical transformation of the MRI target spatial coordinates from prone to supine patient’s position, based on breast 3D shape modelling. This enables the identification of the same target on MRI and US examination (Fig.1-2).
System Evaluation
In this preliminary study, we evaluate the clinical feasibility of BreastNav® and we assess the technology’s accuracy by in-vitro and in-vivo tests.
· In vitro test was performed with a commercially breast phantom (CIRS, model-073) that presents amorphous lesions, mimicking the ultrasonic characteristics of tissues found in an average human breast.
· For in-vivo tests, 2 different investigation scenarios were performed, on a total of 5 patients.
MyLab9 US system (Esaote, Italy) equipped with a novel technology for FI (BreastNav®, MedCom GmbH), based on electromagnetic tracking system a linear probe (L4-15, Esaote, Operating bandwidth: 4-15 MHz), coupled with reusable tracking brackets with sensor mounted were used (Fig. 3).
Procedure
The registration procedure between prone MRI and supine US imaging is based on a 3D Adjustable Breast Model (ABM). The idea is that the breast shape during MRI and US examinations is registered correspondingly to the 3D ABM, whereby then both shapes are inherently registered to each other. Two steps are needed:
a. registration phase between prone MRI imaging dataset and the 3D ABM, based on 5 fixed anatomical superficial points to be set on MRI BI: P1 nipple, P2-P3 median and lateral margins, P4 inframammary fold, P5 parasternal line; the positions of these 5 points are pre-defined on the 3D model (fig.4).
b. registration between the 3D ABM and real-time US on supine patient: 2 sweeps, to describe breast profile, are performed with the US probe directly on patient's breast without any pressure, one horizontal from P2 to P3 points and one vertical from P4 to P5 (Fig. 5). To further improve the registration accuracy, especially in case of hypertrophic breast, it’s possible to furtherly combine the single point acquisition (P1 to P5) and the sweep procedure.
Identifying points and/or acquiring sweeps is completed within a few seconds; after performing the two steps above, a finite element method is used to calculate correspondences and deformations based on points, sweeps and surface information.