Abstract
225
Objectives: We have previously reported the design of the MRC-SPECT-II system based on the inverted-compound-eye (ICE) gamma camera concept to offer an >1% detection efficiency while maintaining a sub-500 µm imaging resolution [1]. One of the key challenges of using the ICE camera for SPECT imaging is whether one could develop an accurate gamma ray response function (GRF) for the ICE-cameras, given its super-complex compound-eye aperture design. In this work, we will discuss (I) a combined experimental and analytical approach for deriving the precise GRF, and (II) an experimental imaging study to demonstrate the feasibility of using the ICE-camera for acquiring high-performance SPECT images. These studies would help to overcome one of the major hurdles for actually implement ICE-cameras for practical SPECT imaging.
Methods: 1) Design of the ICE camera and the MRC-SPECT-II System: the inverted compound-eye (ICE) gamma camera design (Fig.1-E & F) allows for a significantly better tradeoff between FOV, resolution and sensitivity, while maintaining super compact system geometry. Based on ICE camera, we have designed the MRC-SPECT-II system (shown in Fig. 2) that has several attractive features: a) a large number of micro-pinholes cameras could be packed in a high density--the system has 1536 pinholes within a dimension of 75 mm (Trans) by 78 mm (Axial); b) with a large number of cameras viewing the FOV, the system could achieve a peak geometry sensitivity of >1% and rich angular sampling while maintaining sub-mm resolution. To demonstrate the advantages of MRC-SPECT-II system, we have used MRC-SPECT-I [2] as a benchmark (Fig. 3) and carried Monte-Carlo simulation studies, including resolution phantom and beta-amyloid phantom imaging. 2) Experimental and Analytical Gamma Ray Response Function (GRF) Model for the ICE Cameras: (I) the prototype ICE camera consisted of a highly pixelated CdTe detector, a 3D printed aperture with 64 micro-pinholes, and a 4D precision stage system. The detector had 256×256 pixels, each with a size of 100×100 μm2 and was equipped with ASIC that had logic to handle the charge sharing caused by small pixels (Fig. 4-E). (II) Due to complicated geometry and discrepancy between the actual fabricated and designed pinhole profiles, one of critical and challenging issue was modeling the ICE response (Fig.5). We developed a hybrid system modeling scheme that combined analytical model and experimental data (Fig.5-A). First, we measured GRFs of 2500 points that were uniformly and finely distributed (with a grid size of 0.3 mm) around the FOV central plane. Since the measurement was carried in a 2D plane, the 2500 GRFs enabled us physically measure the 64 micro-pinholes’ and detector’s response in high accuracy with feasible time. Then, with the estimated geometry from analytical model and measured GRFs, we interpolated GRFs from the 2D plane to the whole 3D space. (III) To evaluate the hybrid modeling and ICE gamma camera performance, we measured system sensitivity and carried resolution phantom studies.
Results: the single ICE camera could have a detection efficiency of >0.025% with a FOV 10 mm (Fig.6-A); the hybrid system modeling method allowed us to achieve a better accuracy-the GRF based the hybrid model matched measured GRF much better than analytical model (Fig.5-D, E and F) and the maximum error of sensitivity was < 8% (analytical model >30%, shown in fig.5-G); the prototype camera could resolve 0.5 mm hot rods in resolution phantom studies (Fig. 6-Bs vs Fig.6-Cs).
Conclusion: The combined experimental and analytical approach allowed us to derive the precise GRF for complicated micro-pinhole cameras in an ICE module. The experimental imaging study has demonstrated the feasibility of using the ICE-camera for acquiring high-performance SPECT images. These studies could help to overcome one of the major hurdles of implement ICE-cameras for MRC-SPECT-II development. Research Support: 5R21EB018001-02 $$graphic_19E2BBFE-720D-4AA1-911E-3435E8473A30$$