Vincent Keereman

Challenges and current methods for attenuation correction in PET/MR

Quantitative PET imaging requires an attenuation map to correct for attenuation. In stand-alone PET or PET/CT, the attenuation map is usually derived from a transmission scan or CT image, respectively. In PET/MR, these methods will most likely not be used. Therefore, attenuation correction has long been regarded as one of the major challenges in the development of PET/MR. In the past few years, much progress has been made in this field. In this review, the challenges faced in attenuation correction for PET/MR are discussed. Different methods have been proposed to overcome these challenges. An overview of the MR-based (template-based and voxel-based), transmission-based and emission-based methods and the results that have been obtained is provided. Although several methods show promising results, no single method fulfils all of the requirements for the ideal attenuation correction method for PET/MR. Therefore, more work is still necessary in this field. To allow implementation in routine clinical practice, extensive evaluation of the proposed methods is necessary to demonstrate robustness and automation.

The effect of errors in segmented attenuation maps on PET quantification

PURPOSEnAccurate attenuation correction is important for PET quantification. Often, a segmented attenuation map is used, especially in MRI-based attenuation correction. As deriving the attenuation map from MRI images is difficult, different errors can be present in the segmented attenuation map. The goal of this paper is to determine the effect of these errors on quantification.nnnMETHODSnThe authors simulated the digital XCAT phantom using the GATE Monte Carlo simulation framework and a model of the Philips Gemini TF. A whole body scan was simulated, spanning an axial field of view of 70 cm. A total of fifteen lesions were placed in the lung, liver, spine, colon, prostate, and femur. The acquired data were reconstructed with a reference attenuation map and with different attenuation maps that were modified to reflect common segmentation errors. The quantitative difference between reconstructed images was evaluated.nnnRESULTSnSegmentation into five tissue classes, namely cortical bone, spongeous bone, soft tissue, lung, and air yielded errors below 5%. Large errors were caused by ignoring lung tissue (up to 45%) or cortical bone (up to 17%). The interpatient variability of lung attenuation coefficients can lead to errors of 10% and more. Up to 20% tissue misclassification from bone to soft tissue yielded errors below 5%. The same applies for up to 10% misclassification from lung to air.nnnCONCLUSIONSnWhen using a segmented attenuation map, at least five different tissue types should be considered: cortical bone, spongeous bone, soft tissue, lung, and air. Furthermore, the interpatient variability of lung attenuation coefficients should be taken into account. Limited misclassification from bone to soft tissue and from lung to air is acceptable, as these do not lead to relevant errors.

Fast generation of 4D PET-MR data from real dynamic MR acquisitions

We have implemented and evaluated a framework for simulating simultaneous dynamic PET-MR data using the anatomic and dynamic information from real MR acquisitions. PET radiotracer distribution is simulated by assigning typical FDG uptake values to segmented MR images with manually inserted additional virtual lesions. PET projection data and images are simulated using analytic forward projections (including attenuation and Poisson statistics) implemented within the image reconstruction package STIR. PET image reconstructions are also performed with STIR. The simulation is validated with numerical simulation based on Monte Carlo (GATE) which uses more accurate physical modelling, but has 150× slower computation time compared to the analytic method for ten respiratory positions and is 7000× slower when performing multiple realizations. Results are validated in terms of region of interest mean values and coefficients of variation for 65 million coincidences including scattered events. Although some discrepancy is observed, agreement between the two different simulation methods is good given the statistical noise in the data. In particular, the percentage difference of the mean values is 3.1% for tissue, 17% for the lungs and 18% for a small lesion. The utility of the procedure is demonstrated by simulating realistic PET-MR datasets from multiple volunteers with different breathing patterns. The usefulness of the toolkit will be shown for performance investigations of the reconstruction, motion correction and attenuation correction algorithms for dynamic PET-MR data.

DigiPET: sub-millimeter spatial resolution small-animal PET imaging using thin monolithic scintillators

A new preclinical PET system based on dSiPMs, called DigiPET, is presented. The system is based on thin monolithic scintillation crystals and exhibits superior spatial resolution at low-cost compared to systems based on pixelated crystals. Current dedicated small-rodent PET scanners have a spatial resolution in the order of 1xa0mm. Most of them have a large footprint, requiring considerable laboratory space. For rodent brain imaging, a PET scanner with sub-millimeter resolution is desired. To achieve this, crystals with a pixel pitch down to 0.5xa0mm have been used. However, fine pixels are difficult to produce and will render systems expensive. In this work, we present the first results with a high-resolution preclinical PET scanner based on thin monolithic scintillators and a large solid angle. The design is dedicated to rat-brain imaging and therefore has a very compact geometry. Four detectors were placed in a square arrangement with a distance of 34.5xa0mm between two opposing detector modules, defining a field of view (FOV) of 32xa0×xa032xa0×xa032xa0mm(3). Each detector consists of a thin monolithic LYSO crystal of 32xa0×xa032xa0×xa02xa0mm(3)xa0optically coupled to a digital silicon photomultiplier (dSiPM). Event positioning within each detector was obtained using the maximum likelihood estimation (MLE) method. To evaluate the system performance, we measured the energy resolution, coincidence resolving time (CRT), sensitivity and spatial resolution. The image quality was evaluated by acquiring a hot-rod phantom filled with (18)F-FDG and a rat head one hour after an (18)F-FDG injection. The MLE yielded an average intrinsic spatial resolution on the detector of 0.54xa0mm FWHM. We obtained a CRT of 680xa0ps and an energy resolution of 18% FWHM at 511xa0keV. The sensitivity and spatial resolution obtained at the center of the FOV were 6.0 cps kBq(-1)xa0and 0.7xa0mm, respectively. In the reconstructed images of the hot-rod phantom, hot rods down to 0.7xa0mm can be discriminated. In conclusion, a compact PET scanner was built using dSiPM technology and thin monolithic LYSO crystals. Excellent spatial resolution and acceptable sensitivity were demonstrated. Promising results were also obtained in a hot-rod phantom and in rat-brain imaging.

Improvement of attenuation correction in time-of-flight PET/MR imaging with a positron-emitting source.

Quantitative PET imaging relies on accurate attenuation correction. Recently, there has been growing interest in combining state-of-the-art PET systems with MR imaging in a sequential or fully integrated setup. As CT becomes unavailable for these systems, an alternative approach to the CT-based reconstruction of attenuation coefficients (μ values) at 511 keV must be found. Deriving μ values directly from MR images is difficult because MR signals are related to the proton density and relaxation properties of tissue. Therefore, most research groups focus on segmentation or atlas registration techniques. Although studies have shown that these methods provide viable solutions in particular applications, some major drawbacks limit their use in whole-body PET/MR. Previously, we used an annulus-shaped PET transmission source inside the field of view of a PET scanner to measure attenuation coefficients at 511 keV. In this work, we describe the use of this method in studies of patients with the sequential time-of-flight (TOF) PET/MR scanner installed at the Icahn School of Medicine at Mount Sinai, New York, NY. Methods: Five human PET/MR and CT datasets were acquired. The transmission-based attenuation correction method was compared with conventional CT-based attenuation correction and the 3-segment, MR-based attenuation correction available on the TOF PET/MR imaging scanner. Results: The transmission-based method overcame most problems related to the MR-based technique, such as truncation artifacts of the arms, segmentation artifacts in the lungs, and imaging of cortical bone. Additionally, the TOF capabilities of the PET detectors allowed the simultaneous acquisition of transmission and emission data. Compared with the MR-based approach, the transmission-based method provided average improvements in PET quantification of 6.4%, 2.4%, and 18.7% in volumes of interest inside the lung, soft tissue, and bone tissue, respectively. Conclusion: In conclusion, a transmission-based technique with an annulus-shaped transmission source will be more accurate than a conventional MR-based technique for measuring attenuation coefficients at 511 keV in future whole-body PET/MR studies.

Influence of detector pixel size, TOF resolution and DOI on image quality in MR-compatible whole-body PET

The optimization of a whole-body PET system remains a challenging task, as the imaging performance is influenced by a complex interaction of different design parameters. However, it is not always clear which parameters have the largest impact on image quality and are most eligible for optimization. To determine this, we need to be able to assess their influence on image quality. We performed Monte-Carlo simulations of a whole-body PET scanner to predict the influence on image quality of three detector parameters: the TOF resolution, the transverse pixel size and depth-of-interaction (DOI)-correction. The inner diameter of the PET scanner was 65xa0cm, small enough to allow physical integration into a simultaneous PET-MR system. Point sources were used to evaluate the influence of transverse pixel size and DOI-correction on spatial resolution as function of radial distance. To evaluate the influence on contrast recovery and pixel noise a cylindrical phantom of 35xa0cm diameter was used, representing a large patient. The phantom contained multiple hot lesions with 5xa0mm diameter. These lesions were placed at radial distances of 50, 100 and 150xa0mm from the center of the field-of-view, to be able to study the effects at different radial positions. The non-prewhitening (NPW) observer was used for objective analysis of the detectability of the hot lesions in the cylindrical phantom. Based on this analysis the NPW-SNR was used to quantify the relative improvements in image quality due to changes of the variable detector parameters. The image quality of a whole-body PET scanner can be improved significantly by reducing the transverse pixel size from 4 to 2.6xa0mm and improving the TOF resolution from 600 to 400xa0ps and further from 400 to 200xa0ps. Compared to pixel size, the TOF resolution has the larger potential to increase image quality for the simulated phantom. The introduction of two layer DOI-correction only leads to a modest improvement for the spheres at radial distance of 150xa0mm from the center of the transaxial FOV.

Hippocampal Deep Brain Stimulation Reduces Glucose Utilization in the Healthy Rat Brain

PurposeThe effects of deep brain stimulation (DBS) have been studied primarily by cellular studies, which lack the ability to elucidate DBS-related responses on a whole-brain scale. 2-Deoxy-2-[18F]fluoro-d-glucose positron emission tomography ([18F]FDG-PET) reflects changes in neural activity throughout the entire brain volume. The aim of this study was to investigate the whole-brain effect of DBS on the glucose utilization in healthy rats.ProceduresSeven rats were implanted with a DBS electrode in the right hippocampus and injected with [18F]FDG to measure the glucose metabolism during DBS.ResultsAnalysis reveals significant DBS-induced decreases in the glucose metabolism in the bilateral hippocampus and other limbic structures.ConclusionsThis study demonstrates that DBS exhibits not only a local effect around the electrode tip but also in other limbic regions. [18F]FDG-PET studies have the potential to provide better insight into the mechanism of action of DBS by simultaneously observing activity at multiple sites in the brain.

Simulation of dynamic PET data from real MR acquisitions

We have implemented a scheme for simulating realistic dynamic PET data from real MR acquisitions. This toolkit uses a series of MR acquisitions, image registrations and segmentations. PET images are simulated assigning typical values to the segmented images, and manually inserting additional lesions. The data are simulated using analytic forward-projections (including attenuation) with STIR [1], providing a fast and simple simulation toolkit that can be used to efficiently study the performance of different reconstruction and motion and attenuation correction approaches for dynamic PET data.

Functional MRI during Hippocampal Deep Brain Stimulation in the Healthy Rat Brain

Deep Brain Stimulation (DBS) is a promising treatment for neurological and psychiatric disorders. The mechanism of action and the effects of electrical fields administered to the brain by means of an electrode remain to be elucidated. The effects of DBS have been investigated primarily by electrophysiological and neurochemical studies, which lack the ability to investigate DBS-related responses on a whole-brain scale. Visualization of whole-brain effects of DBS requires functional imaging techniques such as functional Magnetic Resonance Imaging (fMRI), which reflects changes in blood oxygen level dependent (BOLD) responses throughout the entire brain volume. In order to visualize BOLD responses induced by DBS, we have developed an MRI-compatible electrode and an acquisition protocol to perform DBS during BOLD fMRI. In this study, we investigate whether DBS during fMRI is valuable to study local and whole-brain effects of hippocampal DBS and to investigate the changes induced by different stimulation intensities. Seven rats were stereotactically implanted with a custom-made MRI-compatible DBS-electrode in the right hippocampus. High frequency Poisson distributed stimulation was applied using a block-design paradigm. Data were processed by means of Independent Component Analysis. Clusters were considered significant when p-values were <0.05 after correction for multiple comparisons. Our data indicate that real-time hippocampal DBS evokes a bilateral BOLD response in hippocampal and other mesolimbic structures, depending on the applied stimulation intensity. We conclude that simultaneous DBS and fMRI can be used to detect local and whole-brain responses to circuit activation with different stimulation intensities, making this technique potentially powerful for exploration of cerebral changes in response to DBS for both preclinical and clinical DBS.

Temperature dependence of APD-based PET scanners

PURPOSEnSolid state detectors such as avalanche photodiodes (APDs) are increasingly being used in PET detectors. One of the disadvantages of APDs is the strong decrease of their gain factor with increasing ambient temperature. The light yield of most scintillation crystals also decreases when ambient temperature is increased. Both effects lead to considerable temperature dependence of the performance of APD-based PET scanners. In this paper, the authors propose a model for this dependence and the performance of the LabPET8 APD-based small animal PET scanner is evaluated at different temperatures.nnnMETHODSnThe model proposes that the effect of increasing temperature on the energy histogram of an APD-based PET scanner is a compression of the histogram along the energy axis. The energy histogram of the LabPET system was acquired at 21u2009°C and 25u2009°C to verify the validity of this model. Using the proposed model, the effect of temperature on system sensitivity was simulated for different detector temperature coefficients and temperatures. Subsequently, the effect of short term and long term temperature changes on the peak sensitivity of the LabPET system was measured. The axial sensitivity profile was measured at 21u2009°C and 24u2009°C following the NEMA NU 4-2008 standard. System spatial resolution was also evaluated. Furthermore, scatter fraction, count losses and random coincidences were evaluated at different temperatures. Image quality was also investigated.nnnRESULTSnAs predicted by the model, the photopeak energy at 25u2009°C is lower than at 21u2009°C with a shift of approximately 6% per °C. Simulations showed that this results in an approximately linear decrease of sensitivity when temperature is increased from 21u2009°C to 24u2009°C and energy thresholds are constant. Experimental evaluation of the peak sensitivity at different temperatures showed a strong linear correlation for short term (2.32 kcps/MBq/°C = 12%/°C, R = -0.95) and long term (1.92 kcps/MBq/°C = 10%/°C , R = -0.96) temperature changes. Count rate evaluation showed that although the total count rate is consistently higher at 21u2009°C than at 24u2009°C for different source activity concentrations, this is mainly due to an increase in scattered and random coincidences. The peak total count rate is 400 kcps at both temperatures but is reached at lower activity at 21u2009°C. The peak true count rate is 138 kcps (at 100 MBq) at 21u2009°C and 180 kcps (at 125 MBq) at 24u2009°C. The peak noise equivalent count rate is also lower at 21u2009°C (70 kcps at 70 MBq) than at 24u2009°C (100 kcps at 100 MBq). At realistic activity levels, the scatter fraction is lower at higher temperatures, but at the cost of a strong decrease in true count rate.nnnCONCLUSIONSnA model was proposed for the temperature dependence of APD-based PET scanners and evaluated using the LabPET small animal PET scanner. System sensitivity and count rate performance are strongly dependent on ambient temperature while system resolution is not. The authors results indicate that it is important to assure stable ambient temperature to obtain reproducible results in imaging studies with APD-based PET scanners.