John Floberg

Dynamic PET Denoising with HYPR Processing

HighlY constrained backPRojection (HYPR) is a promising image-processing strategy with widespread application in time-resolved MRI that is also well suited for PET applications requiring time series data. The HYPR technique involves the creation of a composite image from the entire time series. The individual time frames then provide the basis for weighting matrices of the composite. The signal-to-noise ratio (SNR) of the individual time frames can be dramatically improved using the high SNR of the composite image. In this study, we introduced the modified HYPR algorithm (the HYPR method constraining the backprojections to local regions of interest [HYPR-LR]) for the processing of dynamic PET studies. We demonstrated the performance of HYPR-LR in phantom, small-animal, and human studies using qualitative, semiquantitative, and quantitative comparisons. The results demonstrate that significant improvements in SNR can be realized in the PET time series, particularly for voxel-based analysis, without sacrificing spatial resolution. HYPR-LR processing holds great potential in nuclear medicine imaging for all applications with low SNR in dynamic scans, including for the generation of voxel-based parametric images and visualization of rapid radiotracer uptake and distribution.

Alkylphosphocholine Analogs for Broad-Spectrum Cancer Imaging and Therapy

Tumor-specific alkylphosphocholine analogs were evaluated as imaging and therapy agents in patients and in animal models of human cancer. A Broad View of Cancer Many consider targeted or molecular imaging to be the optimal way to image cancer. Weichert and colleagues feel differently: Uptake of certain small molecules by all cancer cells can give a broad view of cancer, and perhaps also treat it. These small molecules are alkylphosphocholine (APC) analogs, which are taken up preferentially by cancer cells—as compared to, for example, fibroblasts—via plasma membranes and transported into the cells by lipid rafts. The authors tested the uptake of radiolabeled APC analogs in vitro and in vivo in animals in 57 different spontaneous and transgenic tumors, of both human and rodent origin. Because of the well-established efficacy of radiotherapy, the authors demonstrated that the APC analogs could be used to not only visualize tumors but also kill them. Translating this to cancer patients, Weichert et al. showed preliminary preferential uptake of a radiolabeled APC analog in brain tumors. These broadly applicable imaging and therapeutic APC-based agents have been tested in dozens of different human cancers, and preliminarily in people, and are now well poised for further translation to clinical trials. Many solid tumors contain an overabundance of phospholipid ethers relative to normal cells. Capitalizing on this difference, we created cancer-targeted alkylphosphocholine (APC) analogs through structure-activity analyses. Depending on the iodine isotope used, radioiodinated APC analog CLR1404 was used as either a positron emission tomography (PET) imaging (124I) or molecular radiotherapeutic (131I) agent. CLR1404 analogs displayed prolonged tumor-selective retention in 55 in vivo rodent and human cancer and cancer stem cell models. 131I-CLR1404 also displayed efficacy (tumor growth suppression and survival extension) in a wide range of human tumor xenograft models. Human PET/CT (computed tomography) and SPECT (single-photon emission computed tomography)/CT imaging in advanced-cancer patients with 124I-CLR1404 or 131I-CLR1404, respectively, demonstrated selective uptake and prolonged retention in both primary and metastatic malignant tumors. Combined application of these chemically identical APC-based radioisosteres will enable personalized dual modality cancer therapy of using molecular 124I-CLR1404 tumor imaging for planning 131I-CLR1404 therapy.

Surgical decision making in temporal lobe epilepsy: A comparison of [18F]FDG-PET, MRI, and EEG

OBJECTIVES The goals of this work were (1) to determine the effect of [(18)F]fluorodeoxyglucose positron emission tomography (FDG-PET), MRI, and EEG on the decision to perform temporal lobe epilepsy (TLE) surgery, and (2) to determine if FDG-PET, MRI, or EEG predicts surgical outcome. METHODS All PET scans ordered (2000-2010) for epilepsy or seizures were tabulated. Medical records were investigated to determine eligibility and collect data. Statistical analysis included odds ratios, κ statistics, univariate analysis, and logistic regression. RESULTS Of the 186 patients who underwent FDG-PET, 124 had TLE, 50 were surgical candidates, and 34 had surgery with post-operative follow-up. Median length of follow-up was 24 months. MRI, FDG-PET, and EEG were significant predictors of surgical candidacy (P<0.001) with odds ratios of 42.8, 20.4, and 6.3, respectively. FDG-PET was the only significant predictor of postoperative outcome (P<0.01). CONCLUSION MRI showed a trend toward having the most influence on surgical candidacy, but only FDG-PET predicted surgical outcome.

Improved kinetic analysis of dynamic PET data with optimized HYPR-LR

PURPOSE Highly constrained backprojection-local reconstruction (HYPR-LR) has made a dramatic impact on magnetic resonance angiography (MRA) and shows promise for positron emission tomography (PET) because of the improvements in the signal-to-noise ratio (SNR) it provides dynamic images. For PET in particular, HYPR-LR could improve kinetic analysis methods that are sensitive to noise. In this work, the authors closely examine the performance of HYPR-LR in the context of kinetic analysis, they develop an implementation of the algorithm that can be tailored to specific PET imaging tasks to minimize bias and maximize improvement in variance, and they provide a framework for validating the use of HYPR-LR processing for a particular imaging task. METHODS HYPR-LR can introduce errors into non sparse PET studies that might bias kinetic parameter estimates. An implementation of HYPR-LR is proposed that uses multiple temporally summed composite images that are formed based on the kinetics of the tracer being studied (HYPR-LR-MC). The effects of HYPR-LR-MC and of HYPR-LR using a full composite formed with all the frames in the study (HYPR-LR-FC) on the kinetic analysis of Pittsburgh compound-B ([11C]-PIB) are studied. HYPR-LR processing is compared to spatial smoothing. HYPR-LR processing was evaluated using both simulated and human studies. Nondisplaceable binding potential (BP(ND)) parametric images were generated from fifty noise realizations of the same numerical phantom and eight [(11)C]-PIB positive human scans before and after HYPR-LR processing or smoothing using the reference region Logan graphical method and receptor parametric mapping (RPM2). The bias and coefficient of variation in the frontal and parietal cortex in the simulated parametric images were calculated to evaluate the absolute performance of HYPR-LR processing. Bias in the human data was evaluated by comparing parametric image BP(ND) values averaged over large regions of interest (ROIs) to Logan estimates of the BP(ND) from TACs averaged over the same ROIs. Variance was assessed qualitatively in the parametric images and semiquantitatively by studying the correlation between voxel BP(ND) estimates from Logan analysis and RPM2. RESULTS Both the simulated and human data show that HYPR-LR-FC overestimates BP(ND) values in regions of high [(11)C]-PIB uptake. HYPR-LR-MC virtually eliminates this bias. Both implementations of HYPR-LR reduce variance in the parametric images generated with both Logan analysis and RPM2, and HYPR-LR-FC provides a greater reduction in variance. This reduction in variance nearly eliminates the noise-dependent Logan bias. The variance reduction is greater for the Logan method, particularly for HYPR-LR-MC, and the variance in the resulting Logan images is comparable to that in the RPM2 images. HYPR-LR processing compares favorably with spatial smoothing, particularly when the data are analyzed with the Logan method, as it provides a reduction in variance with no loss of spatial resolution. CONCLUSIONS HYPR-LR processing shows significant potential for reducing variance in parametric images, and can eliminate the noise-dependent Logan bias. HYPR-LR-FC processing provides the greatest reduction in variance but introduces a positive bias into the BP(ND) of high-uptake border regions. The proposed method for forming HYPR composite images, HYPR-LR-MC, eliminates this bias at the cost of less variance reduction.

Nonlinear spatio-temporal filtering of dynamic PET data using a four-dimensional Gaussian filter and expectation-maximization deconvolution.

We introduce a method for denoising dynamic PET data, spatio-temporal expectation-maximization (STEM) filtering, that combines four-dimensional Gaussian filtering withEMdeconvolution. The initial Gaussian filter suppresses noise at a broad range of spatial and temporal frequencies and EM deconvolution quickly restores the frequencies most important to the signal. We aim to demonstrate that STEM filtering can improve variance in both individual time frames and in parametric images without introducing significant bias. We evaluate STEM filtering with a dynamic phantom study, and with simulated and human dynamic PET studies of a tracer with reversible binding behaviour, [C-11]raclopride, and a tracer with irreversible binding behaviour, [F-18]FDOPA. STEM filtering is compared to a number of established three and four-dimensional denoising methods. STEM filtering provides substantial improvements in variance in both individual time frames and in parametric images generated with a number of kinetic analysis techniques while introducing little bias. STEM filtering does bias early frames, but this does not affect quantitative parameter estimates. STEM filtering is shown to be superior to the other simple denoising methods studied. STEM filtering is a simple and effective denoising method that could be valuable for a wide range of dynamic PET applications.