Will high-resolution/high-sensitivity SPECT ensure that PET is not the only survivor in nuclear medicine during the next decade?

Abstract

Dear Sir, In a recent issue of the journal, Alavi and Basu [1] on the one hand, and Mariani et al. [2] on the other, debated the future of planar scintigraphy and single photon emission computed tomography (SPECT) in the face of ever better performing positron emission tomography (PET). The pros and cons for the use of single photon emitters, positron emitters and the related tracers were thoroughly discussed. It is, however, our opinion that the discussion relating to the physics and technology of SPECT was biased by the sidestepping of the most recent advances in SPECT, although some points were briefly raised by Mariani et al. [2]. From the beginning, PET was developed as a fully tomographic technique [3]. In the pioneering studies, the detectors, although very limited in number in those early times, were already being placed around the patient’s body. Breakthroughs in crystal technology, electronics, attenuation correction, scatter correction and tomographic reconstruction (Fourier rebinning, 2-D and 3-D iterative reconstruction including resolution and/or time-of-flight consideration) have allowed us to benefit from all the information collected by the detectors. This has led to the modern PET scanners with annular detectors and an axial field-of-view of 15–25 cm with full quantitative 3-D tomographic capabilities [4, 5]. Meanwhile, SPECT continued to be performed with a rotating parallel collimated Anger camera. The only major advances were the dual-head and triple-head cameras, which increased the sensitivity, and the iterative reconstruction algorithms, which improved the overall image quality [6]. Since the beginning of the 1990s, considerable effort has been focused on developing high-resolution SPECT systems by replacing the traditional parallel hole collimators with converging collimators such as fan-beams, cone-beams, cardio-focals, slit-slat collimators and, last but not least, pinholes [7]. Pinhole SPECT was primarily studied in the context of small-animal imaging [8, 9]. In the beginning, a single pinhole mounted on a commercial Anger camera was used [8]. Systems with multiple pinholes with or without the multiplexing of the projections were quickly developed and finally nonrotating systems with a large number of small pinholes came onto the market [9]. Such systems combined high resolution, high sensitivity and collection of the data needed for tomographic reconstruction, without the need to rotate the detectors. This meant that dynamic studies became as easy in SPECT as they were in PET, and with an even better spatial resolution [9]. Meanwhile, a few studies were being devoted to the application of single pinhole SPECT to humans. They demonstrated the feasibility of pinhole SPECT studies of limited volumes of interest (VOI) under not too restrictive conditions: projections could only be collected over 180° around the VOI [10, 11]; the distance from the pinhole aperture to the VOI centre did not necessarily need to be kept constant [12]; the pinhole could be moderately tilted [10–12]; and the detector uniformity requirements were shown to be comparable to those needed for SPECT with conventional parallel hole collimators [13]. Moreover, the introduction of resolution recovery in the iterative reconstruction process was shown to be dramatically effective both in improving the resolution and lowering the noise content of the images [14]. The studies were not limited to the technical feasibility of pinhole SPECT in humans Eur J Nucl Med Mol Imaging (2009) 36:533–535 DOI 10.1007/s00259-008-1026-4