Seungbin Bae

Novel positioning method using Gaussian mixture model for a monolithic scintillator-based detector in positron emission tomography

We developed a high precision position decoding method for a positron emission tomography (PET) detector that consists of a thick slab scintillator coupled with a multichannel photomultiplier tube (PMT). The DETECT2000 simulation package was used to validate light response characteristics for a 48.8 mm×48.8 mm×10 mm slab of lutetium oxy- orthosilicate coupled to a 64 channel PMT. The data are then combined to produce light collection histograms. We employed a Gaussian mixture model (GMM) to parameterize the composite light response with multiple Gaussian mixtures. In the training step, light photons acquired by N PMT channels was used as an N-dimensional feature vector and were fed into a GMM training model to generate optimal parameters for M mixtures. In the positioning step, we decoded the spatial locations of incident photons by evaluating a sample feature vector with respect to the trained mixture parameters. The average spatial resolutions after positioning with four mixtures were 1.1 mm full width at half maximum (FWHM) at the corner and 1.0 mm FWHM at the center section. This indicates that the proposed algorithm achieved high performance in both spatial resolution and posi- tioning bias, especially at the corner section of the detector. C � 2011 Society

FPGA-based multichannel data acquisition system for prototype in-beam PET

In-beam positron emission tomography (PET) is a clinically proven imaging technique for the online investigation of position emitters induced by hadron irradiation. Because the PET involves the use of many photodetectors, the data acquisition (DAQ) system of the PET system requires many elements and a high processing speed to handle many input signals and complex data sets simultaneously. To create a fast and compact DAQ system for a prototype in-beam PET system, we used an FPGA-based multichannel analysis board and developed a firmware program that was geared to acquire the prototype in-beam PET data. As a result, a flood map of the detector block with 22Na 10μCi sources was acquired, and all pixels were well separated. 18F-FDG (4-mm diameter, 3 mCi), which was located in the center of a PET ring, was explored to acquire preliminary results from the prototype in-beam PET system. We produced an energy histogram with an energy resolution of 26% at 511 keV and a reconstructed point image. The measured maximum count rate on the host PC using the developed DAQ system was 120,000 cps. Although there were many improvements in terms of the count rate, the calculation of the pulse timing and correction methods can still be improved, and we were able to assess the feasibility of the prototype in-beam PET system from the achieved preliminary results.

Registered collimation for pixellated SPECT detectors

We investigated a high sensitive collimator (registered collimator) design and associated image reconstruction methods. Its opening was designed to align with the CZT pixel by having same pitch and shape. By the combination of wide opening holes and low septa height, the sensitivity of the collimator can increase dramatically. However, it may also lead to loss of system resolution. The goal of this study is 1) to determine the optimal parameters of the collimator which can increase sensitivity and decrease the charge sharing effect, and 2) to develop system matrix for iterative image reconstruction that can adequately model the proposed collimation scheme to achieve high resolution recovery in spite of using wider collimation holes. Thickness and height of septa in the registered collimator were analytically determined to 1 mm which blocks 97% of 140keV gamma rays and 5.48 mm which indicates 0.2 % efficiency that equivalently means about 10 times higher counts than low energy high resolution(LEHR) collimator. Then 2D fan beam-based geometrical response function(GRF) was developed to model the wide opening holes in the registered collimator. The GRF was included in the system matrix of ordered subset expectation maximization(OSEM) method to compensate the loss of resolution. For performance comparison, GATE simulation was performed using a 5 mm diameter hot sphere phantom, which is located at 1 cm off-centered radial position, and a 40.36 by 40.36 mm CZT detector module. The preliminary results showed that the proposed collimator achieved about 13 times more counts and improved image resolution. In the future, we will keep our experiments with3D cone beam model as well as 2D fan beam model using more realistic phantoms.

New Virtual Frisch-Grid CdZnTe Detector Design With Sub-Millimeter Spatial Resolution

We evaluated the performance of a position-sensitive virtual Frisch-grid (VFG) CdZnTe detector, 6 mm ×6 mm ×15 mm, via sensing strips on its side surfaces. Once the signals were collected from the anode, and from four or eight strips attached to the detectors sides, we assessed the anodes energy spectra and derived histograms from the side electrodes to evaluate the feasibility of achieving sub-millimeter spatial resolution in the X-Y plane. Using a highly collimated 30-keV X-ray beam at the National Synchrotron Light Source, and applying corrections to the raw data, we determined the photon-interaction points by conventional Anger logic and via a more sophisticated statistics-based positioning (SBP) algorithm. With the VFG detectors current configuration, we achieved a resolution below 1 mm, even for low-energy X-rays.

Simulation Study of a Novel Target Oriented SPECT Design Using a Variable Pinhole Collimator

Purpose: In the past decade, demands for organ specific (target oriented) single‐photon emission computed tomography (SPECT) is increasing, and several groups have conducted studies on developing clinical dedicated SPECT with pinhole collimator to improve the spatial resolution. However, acceptance angle of the collimator cannot be adjusted to fit the different ROIs of target objects because the shape of pinhole could not be changed, and the magnifying factor cannot be maximized as the collimator‐to‐detector distance is fixed. Furthermore, those dedicated pinhole SPECTs are typically made for a single purpose and therefore possess a drawback in that it cannot be utilized for any other purpose. In this study, we propose a novel SPECT system using variable pinhole collimator (VP SPECT) whose parameters are flexible. Methods: The proposed variable pinhole collimator is modeled on conventional pinhole by piling several tungsten layers of different apertures. Depending on the combination of the holes in each layer, a variety of hole diameters and acceptance angles of the pinhole can be made. In addition, VP SPECT system allows attaching the collimator to the object as close as possible to maximize the sensitivity and adjust the distance of the pinhole from the scintillation detector to optimize the system resolution for each rotation angle, automatically. For quantitative measurement, we compared the sensitivity and spatial resolution of VP SPECT with those of conventional pinhole SPECT. To determine the possibility of the clinical and preclinical use of proposed system, a digital mouse whole‐body (MOBY) phantom is used for simulating the live mouse model. Results: The result of simulation using ultra‐micro hot spot phantom shows that the sensitivity of the proposed VP SPECT system is about 297% of that of the conventional system. While hot rods of diameter 0.6 mm can be distinguished in the image with the proposed VP SPECT system, 1.2‐mm hot rods are barely discernible in the conventional pinhole SPECT image. According to the result of MOBY phantom simulation, heart walls separated by 3 mm were not distinguished in conventional pinhole SPECT images, but were clearly discerned in VP SPECT images. Conclusions: In this study, we designed a novel pinhole collimator for SPECT and presented preliminary results of target oriented imaging with a simulation study. Currently, we are pursuing strategies to realize the proposed system, with the goal to apply the technology into a high‐sensitivity and high‐resolution preclinical SPECT. Should VP SPECT be applied to the clinical setting, we anticipate a high‐sensitivity, high‐resolution system for applications such as heart dedicated SPECT or related fields.

Development of a prototype SPECT system using a variable pinhole collimator

We designed a novel single-photon emission computed tomography (SPECT) system with a variable pinhole (VP) collimator for obtaining high-sensitivity and high-spatial resolution imaging of a region of interest (ROI). Extending the capabilities of conventional SPECT systems, this novel SPECT system with a VP collimator can adjust its driving parameters, such as a collimator diameter, acceptance angle, and distances between the ROI, collimator, and detector at different rotation angles. Previously, we simulated a VP collimator and demonstrated the superiority of the VP SPECT system in terms of its sensitivity and spatial resolution. Based on these previous results, we developed a prototype SPECT system for performance evaluation of a VP collimator. The collimator was made from 99.9% pure tungsten layers, with 16 holes of different diameters. The area of the CsI(Tl) scintillator was 45.7 × 45.7 mm2 (51 × 51 pixels), with pixel dimensions of 0.7 × 0.7 × 5 mm3. The multi-pixel photon counter was tiled into a 2 × 2 configuration with dimensions of 52 × 52 mm2. The collimator, detector, and gantry of this SPECT system were independently controlled. We acquired initial reconstructed images from a point source (Co-57) and an ultra-micro phantom (Tc-99m). The point source and the phantom were clearly resolved, demonstrating satisfactory detector response, signal processing, and imaging processing capabilities. The 1.35mm phantom was also clearly resolved in the ultra-micro phantom.

Reconstruction of In-beam PET for Carbon therapy with prior-knowledge of carbon beam-track

There are two main artifacts in reconstructed images from in-beam positron emission tomography (PET). Unlike generic PET, in-beam PET uses the annihilation photons that occur during heavy ion therapy. Therefore, the geometry of in-beam PET is not a full ring, but a partial ring that has one or two openings around the rings in order for the hadrons to arrive at the tumor without prevention of detector blocks. This causes truncation in the projection data due to an absence of detector modules in the openings. The other is a ring artifact caused by the gaps between detector modules also found in generic PET. To sum up, in-beam PET has two kinds of gap: openings for hadrons, and gaps between the modules. We acquired three types of simulation results from a PET system: full-ring, C-ring and dual head. In this study, we aim to compensate for the artifacts that come from the two types of gap. In the case of truncation, we propose a method that uses prior knowledge of the location where annihilations occur, and we applied the discrete-cosine transform (DCT) gap-filling method proposed by Tuna et al. for inter-detector gap.

Use of virtual Frisch-grid CdZnTe detectors to attain sub-millimeter spatial resolution

The goal of our study was twofold: To determine the distribution of signals in position-sensitive CdZnTe (CZT)-based virtual Frisch-grid detectors (VFGDs) with side-sensing pads, and to evaluate the feasibility of accurately measuring the X- and Y-coordinates where a photon interaction occurs within a single VFGD module. Accordingly, we collected signals from an anode, and from four or eight sensing pads attached to four sides of a CZT crystal. We assessed the anodes energy spectra and derived histograms from the side electrodes so to evaluate the feasibility of employing VFGDs as imaging devices. Using a highly collimated 30-keV X-ray beam at the National Synchrotron Light Source (NSLS), and applying some corrections to the raw signal data, we found that the signals acquired from one side of the detector were well separated from those measured at the opposite side. We also determined the photon interaction points by conventional Anger logic and via a more sophisticated statistics-based positioning (SBP) algorithm. With the current VFGD configuration, preliminary results showed that our positioning methods could increase the resolution above the intrinsic resolution of the VFGD (6 mm). Using SBP, we achieved a resolution below 1 millimeter for low-energy X- and gamma-rays.

Statistics-based position decoding for a block detector

We are developing a block detector to be used in in-beam PET for hardron therapy, which consists of a discrete scintillator array and four round-type PMTs. To improve positioning performance we applied Gaussian mixture model (GMM)-based positioning algorithm that was previously developed by our group. In order to maximize separability of light distributions among adjacent scintillator pixels and thereby optimize the positioning performance, we used partially segmented block scintillator proposed by Chung et al. In partially segmented block scintillator, length of light reflectors between two adjacent discrete scintillators varies depending on the locations of the scintillators in the array. We simulated 3D crystal array with variable length of reflectors so that we extract best combinations of reflector dimensions in the array. With these optimal values, we showed the performance of our positioning algorithms. The DETECT2000 simulation package was used to model a proposed detector. The designed the detector was made up of 13 × 13 array of 4 × 4 × 20 mm3 LSO blocks. Four sides of each crystal was attached with different length of reflectors. We used 2 × 2 one inch PMTs(22 mm effective area) so that four PMTs can share the lights. In GMM-based positioning algorithm, the response of N detector channels is represented by a feature vector. Then it trains the feature vectors to obtain the optimal parameters of M Gaussian mixtures. In evaluation step, we decoded the spatial locations of incidence photons by evaluating the measured feature vector with respect to the trained mixture parameters. The results showed that the average bias were 0 mm. In addition, most of positions for the 13×13 scintillator block were clearly identified.

Reconstruction of dose distribution in in-beam PET for carbon therapy

There are two main artifacts in reconstructed images from in-beam positron emission tomography(PET). Unlike generic PET, in-beam PET uses the annihilation photons which occur during heavy ion therapy. Therefore, the geometry of in-beam PET is not a full ring, but a partial ring in order for the hardrons to arrive the tumor without penetration of detector blocks. The partial ring, however, causes truncation in projection data, due to an absence of detector modules in the openings. The other is ring artifact caused by the gaps between detector modules which can be also founded in generic PET. In this study, we aim to investigate the effect of gaps in reconstructed images and propose possible solutions to compensate the artifacts. We acquired the data by GATE v6.1 with initial ion energies 170,290, 350AMeV of carbon beams. Each detector module consists of a 13 by 13 LYSO crystal array. The dimension of a crystal was 4mm * 4mm * 20 mm and the radius of inner circle of the gantry was 15cm. In case of truncation error, we proposed to get prior knowledge of the location where annihilations occur. Similar to time-of-flight PET reconstruction, we applied a Gaussian distribution to system matrix, through the width of hardron beams in our back-projection routine. Then expectation maximization (EM) updates were performed iteratively. In case of the latter, to fill the gaps, we used the iterative discrete-cosine transform (DCT) domain method proposed by lJygar Tuna, Sari Peltonen, and lJlla Ruotsalainen. The results show that the proposed method can compensate to some extent the error caused by insufficient angle coverage and we can see the path of hardron beam by proposed method. However, we found other artifacts induced beyond the Bragg peak positions the number of iterations increased. We will improve the proposed algorithm to get more accurate dose distribution.