Diffusion-Weighted Imaging (DWI) for Breast Lesion Characterization: The Olea Medical Perspective and the Utilization of Olea Sphere Software

The Importance of Postprocessing and Data Analysis for Breast Diffusion Weighted Imaging

The role of magnetic resonance imaging (MRI) in breast cancer diagnosis and management is evolving as a supplemental screening tool for high-risk populations. Nowadays, a myriad of magnetic resonance (MR) sequences is applicable to breast imaging, with common protocol consisting in T1-weighted contrast-enhanced and T2-weighted imaging. Dynamic contrast-enhanced (DCE) MRI is highly sensitive, improving detection compared with mammography alone and approaching, with automation techniques, a sensitivity of 89% for all breast cancers. However, specificity remains an issue. Though not yet generalized in the clinical practice, diffusion-weighted imaging (DWI) could improve the diagnostic specificity of DCE (i.e., decrease the false positive rate and consequently reduce the number of unnecessary biopsies). If different MRI sequences offer specific information on lesion morphology and physiology, they could, when combined, guide the operator toward improved lesion identification and characterization. However, the widespread use of breast DWI is limited by its high sensitivity factors that diminish image quality (e.g., residual eddy-currents- and susceptibility-induced distortions). Newer DWI sequences are emerging in complement of standard spin-echo echo-planar DWI, such as readout segmented (RS), multishot, simultaneous multislice (SMS), and zoomed two-dimensional spatially selective radiofrequency (2D RF) excitation pulses, to reduce those effects. Yet, there is very little agreement on the optimal protocol to acquire data and even less on the analysis method to characterize the malignant nature of a breast lesion, tumor aggressiveness, or grade.

This lack of consensus in the clinical community leaves the vendor uncertain as to what characteristics in their software product would be the most applicable for the market. In an effort to address these discrepancies that affect clinicians and vendors alike, the European Society of Breast Imaging (EUSOBI) recently published the first guidelines regarding DWI breast parameters from the consensus of an international group of experts. They include specifications for data acquisition: axial acquisition with at least two b values of 0 and 800 s/mm ² , fat saturation, and a minimal spatial resolution of 2 × 2 mm. Guidelines also recommend to compute standard apparent diffusion coefficient (ADC) maps from these raw diffusion images. Yet there is little information on data analysis besides region-of-interest (ROI) positioning. The determination of the possible malignant nature of a specific lesion based on DWI relies on a combination of observations and metrics based on b -800, ADC maps, and a range of ADC values of the ROI.

Diffusion imaging covers a broad spectrum of techniques, from the mapping of apparent diffusion coefficient (ADC) values using the mono-exponential model to advanced diffusion models such as Intravoxel incoherent motion (IVIM) and Diffusion Kurtosis imaging (DKI) (see Chapters 1 and 8 ) or diffusion tensor imaging (DTI) (see Chapter 9 )

In this chapter, we describe the technical options currently available in the Olea Medical portfolio for the advanced postprocessing of diffusion imaging, compatible with clinical practice expectations in terms of time-effectiveness and clinical impact, in breast lesion assessment.

Breast Diffusion Image Pre- and Postprocessing in Olea Sphere