PEM has two detector heads integrated with planar or curved breast compression paddles, which acquire limited-angle tomographic images from incomplete 3D data obtained with a mildly compressed breast (Fig. 9.1a). The positioning of the breast is similar to that in mammography, and two projections (craniocaudal and mediolateral oblique) are usually obtained. The Flex Solo II scanner (Naviscan, San Diego, USA) is a PEM device that was cleared for marketing by the US Food and Drug Administration (FDA) in 2009 [24] and is currently the most common and commercially available dedicated breast PET system in the United States. Usually, a scanning time for one projection is 7–10 min [25, 34]. The in-plane spatial resolution was reported to be 2.4 ± 0.3 mm full width at half maximum (FWHM) for images reconstructed with a three-dimensional (3D) list-mode maximum likelihood expectation maximization (MLEM) algorithm. On the other hand, the cross plane spatial resolution is low, with a reported FWHM of 8.0–8.2 mm [24, 25], but attenuation and scatter corrections are available. This system is equipped with quantitative metrics to measure quantitative values, called PEM uptake values (PUV) [36]. A biopsy capability is included in this model.

The fully tomographic dedicated PET systems are newer generations of PET devices that acquire complete 3D data from an uncompressed breast, and there are several variations in the detector design. One is a dual- or multi-plate detector rotating around the breast (Fig. 9.1b), such as Clear-PEM developed by the Portuguese consortium under the framework of the Crystal Clear Collaboration at CERN [1] and PEM/PET system developed at West Virginia University [32]. Another type of the fully tomographic dedicated PET systems is a ring-shaped detector encircling the breast (Fig. 9.1c), which includes MAMMI (Oncovision, Valencia, Spain) and Elmammo (Shimadzu, Kyoto, Japan).

In Kyoto University, authors examined a scanner performance of the Elmammo prototype (Shimadzu, Kyoto, Japan) and have been performing human imaging with this system since 2009 [26]. Elmammo has a complete ring-shaped detector, consisting of 36 detector blocks arranged in three contiguous rings with 12 detector modules, which have a transaxial diameter of 185 mm and an axial FOV of 155.5 mm (Fig. 9.2). Each detector block has four layers with a 32 × 32 array of 1.4 × 1.4 × 4.5 mm³ lutetium gadolinium oxyorthosilicate (LGSO) crystals, coupled to a 64-channel position-sensitive photomultiplier tubes (PSPMT). This system has also DOI measurement capability. Elmammo is one of those systems that can achieve the highest spatial resolution among PET systems for human applications. With the Elmammo prototype, minimal FWHMs in the radial, tangential, and axial directions are 1.6, 1.7, and 2.0 mm and 0.8, 0.8, and 0.8 mm, for filtered back projection (FBP) and 3D dynamic row-action maximum likelihood algorithm (DRAMA) reconstructions, respectively [26]. These values are much smaller than the FWHM obtained with whole-body PET or PET/CT systems of approximately 5–7 mm [7, 12]. Elmammo has capabilities of attenuation and scatter corrections and holds a quantitative metrics to obtain standard uptake value (SUV). A breast is usually scanned for 5 min with a patient lying in a prone position. Because the axial FOV of the ring-shaped scanner is designed to be large enough to cover the entire depth of a breast, the whole image is obtained at once without changing the position of the detector. Elmammo is commercially available since 2014 in Japan.