ROC and LROC analyses of the effects of lesion contrast, size, and signal-to-noise ratio on detectability in PET images.

UNLABELLED: Image quality in PET is typically assessed using measures such as contrast recovery, noise variation, and signal-to-noise ratio (SNR). However, these criteria do not directly reflect performance in the clinical use of the images. Lesion detection is a critical task in the clinical interpretation of many PET studies. A receiver operating characteristic (ROC) study is an accepted method for quantitatively evaluating detection performance with respect to factors that influence image quality. ROC and localization ROC (LROC) analyses were conducted to investigate the effects of lesion contrast, SNR, and size on detectability of hot lesions in PET images. METHODS: A thorax phantom was imaged with spheres of 3 sizes simulating lesions (0.45, 1.0, and 1.9 mL). The relative activity in the lesions and the total number of counts acquired were each varied by factors of 2 to ascertain the effects of contrast and SNR, respectively. Measured attenuation correction and a standard reconstruction protocol were used. Three nuclear medicine physicians and 6 medical physicists participated as readers, rating each image and indicating the suspected lesion location. The area under the calculated ROC and LROC curves (Az and Az,LROC) were used as measures of detection performance. RESULTS: Detection performance was shown to increase from virtually random (Az approximately 0.5, Az,LROC approximately 0.2) to superior (Az > 0.9, Az,LROC > 0.9) as lesion contrast was increased by 50% and as lesion SNR was doubled. Detection performance was not seen to vary when comparison was made using image-based measures alone. CONCLUSION: This study quantitatively shows that moderate increases in the image-based measures of lesion contrast and SNR give a relatively large increase in the task-based measure of lesion detection as measured by ROC and LROC analyses. Thus, techniques that give modest increases in lesion contrast or SNR are expected to improve detection. Results will be useful in evaluating improvement in detection for various reconstruction, acquisition, and data analysis methods that enhance contrast or noise performance.

Performance evaluation of microPET: a high-resolution lutetium oxyorthosilicate PET scanner for animal imaging.

UNLABELLED: A new dedicated PET scanner, microPET, was designed and developed at the University of California, Los Angeles, for imaging small laboratory animals. The goal was to provide a compact system with superior spatial resolution at a fraction of the cost of a clinical PET scanner. METHODS: The system uses fiberoptic readout of individually cut lutetium oxyorthosilicate (LSO) crystals to achieve high spatial resolution. Each microPET detector consists of an 8 x 8 array of 2 x 2 x 10-mm LSO scintillation crystals that are coupled to a 64-channel photomultiplier tube by optical fibers. The tomograph consists of 30 detectors in a continuous ring with a 17.2-cm diameter and fields of view (FOVs) of 11.25 cm in the transaxial direction and 1.8 cm in the axial direction. The system has eight crystal rings and no interplane septa. It operates exclusively in the three-dimensional mode and has an electronically controlled bed that is capable of wobbling with a radius of 300 microm. We describe the performance of the tomograph in terms of its spatial, energy and timing resolution, as well as its sensitivity and counting-rate performance. We also illustrate its overall imaging performance with phantom and animal studies that demonstrate the potential applications of this device to biomedical research. RESULTS: Images reconstructed with three-dimensional filtered backprojection show a spatial resolution of 1.8 mm at the center of the FOV (CFOV), which remains <2.5 mm for the central 5 cm of the transaxial FOV. The resulting volumetric resolution of the system is <8 microL. The absolute system sensitivity measured with a 0.74 MBq (20 microCi) 68Ge point source at the CFOV is 5.62 Hz/kBq. The maximum noise equivalent counting rate obtained with a 6.4-cm diameter cylinder spanning the central 56% of the FOV is 10 kcps, whereas the scatter fraction is 37% at the CFOV for an energy window of 250-650 keV and the same diameter cylinder. CONCLUSION: This is the first PET scanner to use the new scintillator LSO and uses a novel detector design to achieve high volumetric spatial resolution. The combination of imaging characteristics of this prototype system (resolution, sensitivity, counting-rate performance and scatter fraction) opens up new possibilities in the study of animal models with PET.

ROC and localization ROC analyses of lesion detection in whole-body FDG PET: effects of acquisition mode, attenuation correction and reconstruction algorithm.

UNLABELLED: Receiver operating characteristic (ROC) and localization ROC (LROC) studies were performed to compare lesion detection at the borderline of detectability on images reconstructed with two-dimensional filtered backprojection (FBP) without attenuation correction (a common clinical protocol), three-dimensional FBP without attenuation correction, two-dimensional FBP with segmented attenuation correction and a two-dimensional iterative maximum a posteriori (MAP) algorithm using attenuation correction. Lung cancer was the model for the study because of the prominent role of 18F-fluorodeoxyglucose PET in the staging of lung cancer and the importance of lesion detection for staging. METHODS: Simulated lung cancer lesions were added to two-dimensional and three-dimensional PET data from healthy volunteers. Data were reconstructed using the four methods. Four nuclear medicine physicians evaluated the images. Detection performance with each method was compared using ROC and LROC analysis. Jackknife analysis provided estimates of statistical significance for differences across all readers for the ROC results. RESULTS: ROC and LROC results indicated statistically significant degradation in detection performance with three-dimensional acquisition (average area under ROC curves [Az] 0.51; average area under LROC curves [A(z,LROC)] 0.13) and segmented attenuation correction (average Az 0.59; average Az,LROC 0.29) compared with two-dimensional FBP without attenuation correction (average Az 0.79; average A(z,LROC) 0.54). ROC and LROC results indicated an improvement in detection performance with iterative MAP reconstruction (average Az 0.83; average A(z,LROC) 0.64) compared with two-dimensional FBP reconstruction; this improvement was not statistically significant. CONCLUSION: Use of segmented attenuation correction or three-dimensional acquisition with FBP reconstruction is not expected to improve detection of lung lesions on whole-body PET images compared with images with two-dimensional FBP without attenuation correction. The potential improvement in detection obtained with an iterative MAP reconstruction method is small compared with that obtained with two-dimensional FBP without attenuation correction.

High-resolution 3D Bayesian image reconstruction using the microPET small-animal scanner.

A Bayesian method is described for reconstruction of high-resolution 3D images from the microPET small-animal scanner. Resolution recovery is achieved by explicitly modelling the depth dependent geometric sensitivity for each voxel in combination with an accurate detector response model that includes factors due to photon pair non-collinearity and inter-crystal scatter and penetration. To reduce storage and computational costs we use a factored matrix in which the detector response is modelled using a sinogram blurring kernel. Maximum a posteriori (MAP) images are reconstructed using this model in combination with a Poisson likelihood function and a Gibbs prior on the image. Reconstructions obtained from point source data using the accurate system model demonstrate a potential for near-isotropic FWHM resolution of approximately 1.2 mm at the center of the field of view compared with approximately 2 mm when using an analytic 3D reprojection (3DRP) method with a ramp filter. These results also show the ability of the accurate system model to compensate for resolution loss due to crystal penetration producing nearly constant radial FWHM resolution of 1 mm out to a 4 mm radius. Studies with a point source in a uniform cylinder indicate that as the resolution of the image is reduced to control noise propagation the resolution obtained using the accurate system model is superior to that obtained using 3DRP at matched background noise levels. Additional studies using pie phantoms with hot and cold cylinders of diameter 1-2.5 mm and 18FDG animal studies appear to confirm this observation.

An evaluation of exact and approximate 3-D reconstruction algorithms for a high-resolution, small-animal PET scanner.

MicroPET is a low-cost, high-resolution positron emission tomography (PET) scanner designed for imaging small animals. MicroPET operates exclusively without septa, acquiring fully three-dimensional (3-D) data sets. The performance of the projection-reprojection (3DRP), variable axial rebinning (VARB), single slice rebinning (SSRB), and Fourier rebinning (FORE) methods for reconstruction of microPET data were evaluated. The algorithms were compared with respect to resolution, noise variance, and reconstruction time. Results suggested that the 3DRP algorithm gives the best combination of resolution and noise performance in 9 min of reconstruction time on a Sun UltraSparc I workstation. The FORE algorithm provided the most acceptable accelerated method of reconstruction, giving similar resolution performance with a 10%-20% degradation in noise variance in under 2 min. Significant degradation in the axial resolution was measured with the VARB and SSRB methods, offsetting the decrease in reconstruction time achieved with those methods. In-plane angular mashing of the 3-D data before reconstruction led to a 50% reduction in reconstruction time but also introduced unacceptable tangential blurring artifacts. This thorough evaluation of analytical 3-D reconstruction techniques allowed for optimal selection of a reconstruction method for the diverse range of microPET applications.