Design study of a high-resolution and ultrahigh-sensitivity brain SPECT system for imaging medically intractable epilepsy

Objectives: SPECT is a clinical imaging modality widely used to assess dynamic changes in cerebral perfusion before, during, and after a seizure. Ictal SPECT is routinely used to reveal an area of hyperperfusion in the epileptogenic region, surrounded by areas of hypoperfusion, which provides valuable information for the localization of the epileptic network [1, 2]. These clinical studies could benefit from future SPECT instruments with a greatly improved sensitivity and completely stationary image acquisition geometry. This would significantly improve the quantitation and allow for rapid SPECT acquisitions at multiple time points to capture ictal cerebral blood flow and post-ictal propagation. In this work, we present the results from a design study of a high-performance brain SPECT system, a compact and stationary SPECT system based on pixelated CZT detectors [3], a newly developed multi-detector readout circuitry [4], and a spherical dense camera array (DCA) design equipped with micro-slit and micro-ring apertures [5]. The unique design of the DCA-SPECT system could greatly improve the seizure localization yield for patients with medically intractable epilepsies undergoing presurgical evaluation.

Methods: The system design concept is based on the DCA design (Fig.1A) [5] where a large number (~ 500) of independent micro-camera elements (MCE) are closely packed in a semi-spherical surface surrounding the patient’s head. Each MCE consists of a high-resolution CZT detector coupled to a pinhole, micro-ring or micro-slit aperture in a highly de-magnifying geometry (Fig.1B). The DCA design combines some attractive features: the large number of different cameras surrounding the object ensures an improved sensitivity, maintains an excellent imaging resolution, and shows an unprecedented density of angular sampling in a clinically relevant FOV, which is usually lacking in stationary SPECT systems. Additionally, the highly de-magnifying geometry would require an overall reduced detector volume in comparison to a conventional magnifying geometry, and therefore a lowered hardware cost. The CZT detection unit (Fig.1C) consists of a 5-mm thick CZT crystal, 20 mm × 20 mm in size, and bump-bonded to a HEXITEC ASIC with 80 × 80 pixels of 250 μm ×250 μm pitch [3]. The proposed detector is characterized by an excellent energy resolution (0.76% ± 0.14 FWHM at 140 keV), that makes the system well-suited for multi-isotope SPECT imaging (Fig.1D). The system will be equipped with a modular and reconfigurable external readout electronics [4] that we have recently developed for high-speed and ultra-high energy resolution single-photon imaging applications. In the design, we have explored the use of conventional pinholes, as well as non-conventional apertures, such as micro-slits or micro-rings, and compared their performance.

Results: The preliminary results from simulation studies show that the proposed brain SPECT system can achieve a system peak sensitivity ranging from 0.11% when 1-mm D pinholes are used, to 1.38% when micro-rings having a 250-μm annulus opening are used (Fig.1E), while preserving a clinically relevant FOV of 20 cm in diameter and a spatial resolution of 4 mm (Fig.1F). Additionally, we have simulated a realistic ictal brain perfusion phantom [6] to show the improved imaging performance offered by the DCA-SPECT system for imaging ictal and post-ictal brain perfusion patterns (Fig.1G).

Conclusions: Several commercial clinical SPECT systems based on solid-state detectors have been available for several years [7]. The proposed SPECT system design brings together the latest CZT detector technology, the novel DCA gamma camera design, and non-conventional micro-slit and micro-ring apertures into a potential next generation of brain SPECT scanners. The system could offer a dramatically improved imaging performance and potentially bring a radical change in managing epilepsy in clinical practice.