Search

Chest x-ray (CXR) is one of the most commonly used imaging techniques for the detection and diagnosis of pulmonary diseases. One critical component in many computer-aided systems, for either detection or diagnosis in digital CXR, is the accurate segmentation of the lung. Due to low-intensity contrast around lung boundary and large inter-subject variance, it has been challenging to segment lung from structural CXR images accurately. In this work, we propose an automatic Hybrid Segmentation Network (H-SegNet) for lung segmentation on CXR. The proposed H-SegNet consists of two key steps: (1) an image preprocessing step based on a deep learning model to automatically extract coarse lung contours; (2) a refinement step to fine-tune the coarse segmentation results based on an improved principal curve-based method coupled with an improved machine learning method. Experimental results on several public datasets show that the proposed method achieves superior segmentation results in lung CXRs, compared with several state-of-the-art methods., (© 2022 Institute of Physics and Engineering in Medicine.)

Deep convolutional neural network (DCNN) has shown great success in various medical image segmentation tasks, including organ-at-risk (OAR) segmentation from computed tomography (CT) images. However, most studies use the dataset from the same source(s) for training and testing so that the ability of a trained DCNN to generalize to a different dataset is not well studied, as well as the strategy to address the issue of performance drop on a different dataset. In this study we investigated the performance of a well-trained DCNN model from a public dataset for thoracic OAR segmentation on a local dataset and explored the systematic differences between the datasets. We observed that a subtle shift of organs inside patient body due to the abdominal compression technique during image acquisition caused significantly worse performance on the local dataset. Furthermore, we developed an optimal strategy via incorporating different numbers of new cases from the local institution and using transfer learning to improve the accuracy and robustness of the trained DCNN model. We found that by adding as few as 10 cases from the local institution, the performance can reach the same level as in the original dataset. With transfer learning, the training time can be significantly shortened with slightly worse performance for heart segmentation.

Dual-energy CT, as well as spectral CT, has a great potential in material decomposition. However, dual-energy CT is difficult to apply to multi-material decomposition because the number of energy bins is limited to two. Current spectral CT systems have more energy bins, but the statistical noise in each energy bin is high because of the decreased photon number, which causes errors in the material decomposition results. In this paper, we propose a dynamic-dual-energy spectral CT for accurate multi-material decomposition. In the course of scanning, the energy threshold of the dynamic-dual-energy detector randomly changes to obtain the spectral information of photons. With the proposed statistical noise-weighted tPRISM algorithm, the multi-energy image reconstruction using dynamic-dual-energy CT data was implemented, followed by multi-material decomposition. Both simulation and experiment results show that the multi-energy reconstruction and multi-material decomposition using the dynamic-dual-energy method are more accurate and have less noise compared with that of the conventional static-multi-energy method with the same number of energy bins. The ring artifacts which are severe in the experimental data simulation and experiment results using the conventional spectral CT method are reduced in great extent when using our proposed method. In conclusion, our proposed dynamic-dual-energy spectral CT method is highly feasible and has a great potential in high-quality multi-material decomposition.

Dual-energy chest radiography (DECR) is a medical imaging technology that can improve diagnostic accuracy. This technique can decompose single-energy chest radiography (SECR) images into separate bone- and soft tissue-only images. This can, however, double the radiation exposure to the patient. To address this limitation, we developed an algorithm for the synthesis of DECR from a SECR through deep learning. To predict high resolution images, we developed a novel deep learning architecture by modifying a conventional U-net to take advantage of the high frequency-dominant information that propagates from the encoding part to the decoding part. In addition, we used the anticorrelated relationship (ACR) of DECR for improving the quality of the predicted images. For training data, 300 pairs of SECR and their corresponding DECR images were used. To test the trained model, 50 DECR images from Yonsei University Severance Hospital and 662 publicly accessible SECRs were used. To evaluate the performance of the proposed method, we compared DECR and predicted images using a structural similarity approach (SSIM). In addition, we quantitatively evaluated image quality calculating the modulation transfer function and coefficient of variation. The proposed model selectively predicted the bone- and soft tissue-only CR images from an SECR image. The strategy for improving the spatial resolution by ACR was effective. Quantitative evaluation showed that the proposed method with ACR showed relatively high SSIM (over 0.85). In addition, predicted images with the proposed ACR model achieved better image quality measures than those of U-net. In conclusion, the proposed method can obtain high-quality bone- and soft tissue-only CR images without the need for additional hardware for double x-ray exposures in clinical practice.

Toxicity to cardiac and coronary structures is an important late morbidity for patients undergoing left-sided breast radiotherapy. Many current studies have relied on estimates of cardiac doses assuming standardised anatomy, with a calculated increase in relative risk of 7.4% per Gy (mean heart dose). To provide individualised estimates for dose, delineation of various cardiac structures on patient images is required. Automatic multi-atlas based segmentation can provide a consistent, robust solution, however there are challenges to this method. We are aiming to develop and validate a cardiac atlas and segmentation framework, with a focus on the limitations and uncertainties in the process. We present a probabilistic approach to segmentation, which provides a simple method to incorporate inter-observer variation, as well as a useful tool for evaluating the accuracy and sources of error in segmentation. A dataset consisting of 20 planning computed tomography (CT) images of Australian breast cancer patients with delineations of 17 structures (including whole heart, four chambers, coronary arteries and valves) was manually contoured by three independent observers, following a protocol based on a published reference atlas, with verification by a cardiologist. To develop and validate the segmentation framework a leave-one-out cross-validation strategy was implemented. Performance of the automatic segmentations was evaluated relative to inter-observer variability in manually-derived contours; measures of volume and surface accuracy (Dice similarity coefficient (DSC) and mean absolute surface distance (MASD), respectively) were used to compare automatic segmentation to the consensus segmentation from manual contours. For the whole heart, the resulting segmentation achieved a DSC of [Formula: see text], with a MASD of [Formula: see text] mm. Quantitative results, together with the analysis of probabilistic labelling, indicate the feasibility of accurate and consistent segmentation of larger structures, whereas this is not the case for many smaller structures, where a major limitation in segmentation accuracy is the inter-observer variability in manual contouring.

PET images commonly suffer from the high noise level and poor signal-to-noise ratio (SNR), thus adversely impacting lesion detectability and quantitative accuracy. In this work, a novel hybrid dual-domain PET denoising approach is proposed, which combines the advantages of both spatial and transform domain filtering to preserve image textures while minimizing quantification uncertainty. Spatial domain denoising techniques excel at preserving high-contrast patterns compared to transform domain filters, which perform well in recovering low-contrast details normally smoothed out by spatial domain filters. For spatial domain filtering, the non-local mean algorithm was chosen owing to its performance in denoising high-contrast features whereas multi-scale curvelet denoising was exploited for the transform domain owing to its capability to recover small details. The proposed hybrid method was compared to conventional post-reconstruction Gaussian and edge preserving bilateral filters. Computer simulations of a thorax phantom containing three small lesions, experimental measurements using the Jaszczak phantom and clinical whole-body PET/CT studies were used to evaluate the performance of the proposed PET denoising technique. The proposed hybrid filter increased the SNR from 8.0 (non-filtered PET image) to 39.3 for small lesions in the computerized thorax phantom, while Gaussian and bilateral filtering led to SNRs of 23.3 and 24.4, respectively. For the experimental Jaszczak phantom, the contrast-to-noise ratio (CNR) improved from 10.84 when using Gaussian smoothing to 14.02 and 19.39 using the bilateral and the proposed hybrid filters, respectively. The clinical studies further demonstrated the superior performance of the hybrid method, yielding a quantification change (the original noisy OSEM image was used as reference in the absence of ground truth) in malignant lesions of  -2.4% compared to  -11.9% and  -6.6% achieved using Gaussian and bilateral filters, respectively. In some cases, the visual difference between the bilateral and hybrid filtered images is not substantial; however the improved CNR score from 11.3 by OSEM to 17.1 and 21.8 by bilateral to the hybrid filtering, respectively, demonstrates the overall gain achieved by the hybrid approach. The proposed hybrid algorithm improved the contrast, SNR and quantitative accuracy compared to Gaussian and bilateral approaches, and can be utilized as an alternative post-reconstruction filter in clinical PET/CT imaging.

We present an ultrasound-driven 4D magnetic resonance imaging (US-4DMRI) method for respiratory motion imaging in the thorax and abdomen. The proposed US-4DMRI comes along with a high temporal resolution, and allows for organ motion imaging beyond a single respiratory cycle. With the availability of the US surrogate both inside and outside the MR bore, 4D MR images can be reconstructed for 4D treatment planning and online respiratory motion prediction during radiotherapy. US-4DMRI relies on simultaneously acquired 2D liver US images and abdominal 2D MR multi-slice scans under free respiration. MR volumes are retrospectively composed by grouping the MR slices corresponding to the most similar US images. We present two different US similarity metrics: an intensity-based approach, and a similarity measure relying on predefined fiducials which are being tracked over time. The proposed method is demonstrated on MR liver scans of eight volunteers acquired over a duration of 5.5 min each at a temporal resolution of 2.6 Hz with synchronous US imaging at 14 Hz-17 Hz. Visual inspection of the reconstructed MR volumes revealed satisfactory results in terms of continuity in organ boundaries and blood vessels. In quantitative leave-one-out experiments, both US similarity metrics reach the performance level of state-of-the-art navigator-based approaches.

The 'simultaneous maximum-likelihood attenuation correction factors' (sMLACF) algorithm presented here, is an iterative algorithm to calculate the maximum-likelihood estimate of the activity λ and the attenuation factors a in time-of-flight positron emission tomography, and this from emission data only. Hence sMLACF is an alternative to the MLACF algorithm. sMLACF is derived using the generalized expectation-maximization principle by introducing an appropriate set of complete data. The resulting iteration step yields a simultaneous update of λ and a which, in addition, enforces in a natural way the constraints [Formula: see text] where [Formula: see text] is a fixed lower bound that ensures the boundedness of the reconstructed activities. Some properties-like the monotonic increase of the likelihood and the asymptotic regularity of the estimated [Formula: see text]-of sMLACF are proven. Comparison of sMLACF with MLACF for two data sets reveals that both algorithms show very similar results, although sMLACF converges slower.

A novel breathing phantom was designed for being used in conventional and ion-beam radiotherapy as well as for medical imaging. Accurate dose delivery and patient safety are aimed to be verified for four-dimensional (4D) treatment techniques compensating for breathing-induced tumor motion. The phantom includes anthropomorphic components representing an average human thorax. It consists of real tissue equivalent materials to fulfill the requirements for dosimetric experiments and imaging purposes. The different parts of the torso (lungs, chest wall, and ribs) and the tumor can move independently. Simple regular movements, as well as more advanced patient-specific breathing cycles are feasible while a reproducible setup can be guaranteed. The phantom provides the flexibility to use different types of dosimetric devices and was designed in a way that it is robust, transportable and easy to handle. Tolerance levels and the reliability of the phantom setup were determined in combination with tests on motion accuracy and reproducibility by using infrared optical tracking technology. Different imaging was performed including positron emission tomography imaging, 4D computed tomography as well as real-time in-room imaging. The initial dosimetric benchmarking studies were performed in a photon beam where dose parameters are predictable and the dosimetric procedures well established.

Motion during data acquisition is a known source of error in medical tomography, resulting in blur artefacts in the regions that move. It is critical to reduce these artefacts in applications such as image-guided radiation therapy as a clearer image translates into a more accurate treatment and the sparing of healthy tissue close to a tumour site. Most research in 4D x-ray tomography involving the thorax relies on respiratory phase binning of the acquired data and reconstructing each of a set of images using the limited subset of data per phase. In this work, we demonstrate a motion-compensation method to reconstruct images from the complete dataset taken during breathing without recourse to phase-binning or breath-hold techniques. As long as the motion is sufficiently well known, the new method can accurately reconstruct an image at any time during the acquisition time span. It can be applied to any iterative reconstruction algorithm.

Excessive radiation exposure is still a major concern in 4D cone-beam computed tomography (4D-CBCT) due to its prolonged scanning duration. Radiation dose can be effectively reduced by either under-sampling the x-ray projections or reducing the x-ray flux. However, 4D-CBCT reconstruction under such low-dose protocols is prone to image artifacts and noise. In this work, we propose a novel joint regularization-based iterative reconstruction method for low-dose 4D-CBCT. To tackle the under-sampling problem, we employ spatiotemporal tensor framelet (STF) regularization to take advantage of the spatiotemporal coherence of the patient anatomy in 4D images. To simultaneously suppress the image noise caused by photon starvation, we also incorporate spatiotemporal nonlocal total variation (SNTV) regularization to make use of the nonlocal self-recursiveness of anatomical structures in the spatial and temporal domains. Under the joint STF-SNTV regularization, the proposed iterative reconstruction approach is evaluated first using two digital phantoms and then using physical experiment data in the low-dose context of both under-sampled and noisy projections. Compared with existing approaches via either STF or SNTV regularization alone, the presented hybrid approach achieves improved image quality, and is particularly effective for the reconstruction of low-dose 4D-CBCT data that are not only sparse but noisy.

Motion-induced range changes and incorrectly placed dose spots strongly affect the quality of pencil-beam-scanned (PBS) proton therapy, especially in thoracic tumour sites, where density changes are large. Thus motion-mitigation techniques are necessary, which must be validated in a realistic patient-like geometry. We report on the development and characterisation of a dynamic, anthropomorphic, thorax phantom that can realistically mimic thoracic motions and anatomical features for verifications of proton and photon 4D treatments. The presented phantom is of an average thorax size, and consists of inflatable, deformable lungs surrounded by a skeleton and skin. A mobile 'tumour' is embedded in the lungs in which dosimetry devices (such as radiochromic films) can be inserted. Motion of the tumour and deformation of the thorax is controlled via a custom made pump system driving air into and out of the lungs. Comprehensive commissioning tests have been performed to evaluate the mechanical performance of the phantom, its visibility on CT and MR imaging and its feasibility for dosimetric validation of 4D proton treatments. The phantom performed well on both regular and irregular pre-programmed breathing curves, reaching peak-to-peak amplitudes in the tumour of  <20 mm. Some hysteresis in the inflation versus deflation phases was seen. All materials were clearly visualised in CT scans, and all, except the bone and lung components, were MRI visible. Radiochromic film measurements in the phantom showed that imaging for repositioning was required (as for a patient treatment). Dosimetry was feasible with Gamma Index agreements (4%/4 mm) between film dose and planned dose  >90% in the central planes of the target. The results of this study demonstrate that this anthropomorphic thorax phantom is suitable for imaging and dosimetric studies in a thoracic geometry closely-matched to lung cancer patients under realistic motion conditions.

Both in radiography and computed tomography (CT), recently emerged energy-resolved x-ray photon counting detectors enable the identification and quantification of individual materials comprising the inspected object. However, the approaches used for these operations require highly accurate x-ray images. The accuracy of the images is severely compromised by the presence of scattered radiation, which leads to a loss of spatial contrast and, more importantly, a bias in radiographic material imaging and artefacts in CT. The aim of the present study was to experimentally evaluate a recently introduced partial attenuation spectral scatter separation approach (PASSSA) adapted for multi-energy imaging. For this purpose, a prototype x-ray system was used. Several radiographic acquisitions of an anthropomorphic thorax phantom were performed. Reference primary images were obtained via the beam-stop (BS) approach. The attenuation images acquired from PASSSA-corrected data showed a substantial increase in local contrast and internal structure contour visibility when compared to uncorrected images. A substantial reduction of scatter induced bias was also achieved. Quantitatively, the developed method proved to be in relatively good agreement with the BS data. The application of the proposed scatter correction technique lowered the initial normalized root-mean-square error (NRMSE) of 45% between the uncorrected total and the reference primary spectral images by a factor of 9, thus reducing it to around 5%.

Image registration is part of a large variety of medical applications including diagnosis, monitoring disease progression and/or treatment effectiveness and, more recently, therapy guidance. Such applications usually involve several imaging modalities such as ultrasound, computed tomography, positron emission tomography, x-ray or magnetic resonance imaging, either separately or combined. In the current work, we propose a non-rigid multi-modal registration method (namely EVolution: an edge-based variational method for non-rigid multi-modal image registration) that aims at maximizing edge alignment between the images being registered. The proposed algorithm requires only contrasts between physiological tissues, preferably present in both image modalities, and assumes deformable/elastic tissues. Given both is shown to be well suitable for non-rigid co-registration across different image types/contrasts (T1/T2) as well as different modalities (CT/MRI). This is achieved using a variational scheme that provides a fast algorithm with a low number of control parameters. Results obtained on an annotated CT data set were comparable to the ones provided by state-of-the-art multi-modal image registration algorithms, for all tested experimental conditions (image pre-filtering, image intensity variation, noise perturbation). Moreover, we demonstrate that, compared to existing approaches, our method possesses increased robustness to transient structures (i.e. that are only present in some of the images).

A comprehensive artefact correction method for clinical cone beam CT (CBCT) images acquired for image guided radiation therapy (IGRT) on a commercial system is presented. The method is demonstrated to reduce artefacts and recover CT-like Hounsfield units (HU) in reconstructed CBCT images of five lung cancer patients. Projection image based artefact corrections of image lag, detector scatter, body scatter and beam hardening are described and applied to CBCT images of five lung cancer patients. Image quality is evaluated through visual appearance of the reconstructed images, HU-correspondence with the planning CT images, and total volume HU error. Artefacts are reduced and CT-like HUs are recovered in the artefact corrected CBCT images. Visual inspection confirms that artefacts are indeed suppressed by the proposed method, and the HU root mean square difference between reconstructed CBCTs and the reference CT images are reduced by 31% when using the artefact corrections compared to the standard clinical CBCT reconstruction. A versatile artefact correction method for clinical CBCT images acquired for IGRT has been developed. HU values are recovered in the corrected CBCT images. The proposed method relies on post processing of clinical projection images, and does not require patient specific optimisation. It is thus a powerful tool for image quality improvement of large numbers of CBCT images.

Four dimensional cone beam computed tomography (4DCBCT) is an image guidance strategy used for patient positioning in radiotherapy. In conventional implementations of 4DCBCT, a constant gantry speed and a constant projection pulse rate are used. Unfortunately, this leads to higher imaging doses than are necessary because a large number of redundant projections are acquired. In theoretical studies, we have previously demonstrated that by suppressing redundant projections the imaging dose can be reduced by 40-50% for a majority of patients with little reduction in image quality. The aim of this study was to experimentally realise the projection suppression technique, which we have called Respiratory Triggered 4DCBCT (RT-4DCBCT). A real-time control system was developed that takes the respiratory signal as input and computes whether to acquire, or suppress, the next projection trigger during 4DCBCT acquisition. The CIRS dynamic thorax phantom was programmed with a 2 cm peak-to-peak motion and periods ranging from 2 to 8 s. Image quality was assessed by computing the edge response width of a 3 cm imaging insert placed in the phantom as well as the signal to noise ratio of the phantoms tissue and the contrast to noise ratio between the phantoms lung and tissue. The standard deviation in the superior-inferior direction of the 3 cm imaging insert was used to assess intra-phase bin displacement variations with a higher standard deviation implying more motion blur. The 4DCBCT imaging dose was reduced by 8.6%, 41%, 54%, 70% and 77% for patients with 2, 3, 4, 6 and 8 s breathing periods respectively when compared to conventional 4DCBCT. The standard deviation of the intra-phase bin displacement variation of the 3 cm imaging insert was reduced by between 13% and 43% indicating a more consistent position for the projections within respiratory phases. For the 4 s breathing period, the edge response width was reduced by 39% (0.8 mm) with only a 6-7% decrease in the signal to noise and contrast to noise ratios. RT-4DCBCT has been experimentally realised and reduced to practice on a linear accelerator with a measurable imaging dose reductions over conventional 4DCBCT and little degradation in image quality.

Accurate PET quantitation requires a matched attenuation map. Obtaining matched CT attenuation maps in the thorax is difficult due to the respiratory cycle which causes both motion and density changes. Unlike with motion, little attention has been given to the effects of density changes in the lung on PET quantitation. This work aims to explore the extent of the errors caused by pulmonary density attenuation map mismatch on dynamic and static parameter estimates. Dynamic XCAT phantoms were utilised using clinically relevant (18)F-FDG and (18)F-FMISO time activity curves for all organs within the thorax to estimate the expected parameter errors. The simulations were then validated with PET data from 5 patients suffering from idiopathic pulmonary fibrosis who underwent PET/Cine-CT. The PET data were reconstructed with three gates obtained from the Cine-CT and the average Cine-CT. The lung TACs clearly displayed differences between true and measured curves with error depending on global activity distribution at the time of measurement. The density errors from using a mismatched attenuation map were found to have a considerable impact on PET quantitative accuracy. Maximum errors due to density mismatch were found to be as high as 25% in the XCAT simulation. Differences in patient derived kinetic parameter estimates and static concentration between the extreme gates were found to be as high as 31% and 14%, respectively. Overall our results show that respiratory associated density errors in the attenuation map affect quantitation throughout the lung, not just regions near boundaries. The extent of this error is dependent on the activity distribution in the thorax and hence on the tracer and time of acquisition. Consequently there may be a significant impact on estimated kinetic parameters throughout the lung.

Respiratory and cardiac motions are inevitable during the relatively long acquisition time of cardiac positron emission tomography (PET) scan. The correction of the resultant motion blur has become a significant challenge due to recent spatial resolution improvement of the PET scanners. The majority of current motion compensation algorithms are based on gating as a primary step. A new approach based on temporal basis functions is developed to correct respiratory and cardiac motion simultaneously in cardiac PET within the normal scanning time (NAMC-CPET). Simulation and experimental studies are conducted to evaluate and validate the final outputs in comparison to the existing gating methods. A dynamic digital phantom is used to simulate realistic human thorax and abdomen with respiratory and cardiac motions. GATE simulation was run at China National Grid Center to obtain realistic PET data in a reasonable time. Moreover, Tibet minipig experiments were conducted using a preclinical small animal PET scanner developed at HUST to validate the performance of the NAMC-CPET in real data. The results reveal that NAMC-CPET outperformed the existing gating methods (respiratory, cardiac, and dual) in cardiac imaging in term of noise reduction and contrast, especially in short acquisition duration. NAMC-CPET obtained better results in the conducted experiments in terms of contrast and the visibility of the heart. In contrast, the dual gating failed to obtain valuable images in the normal scan time due to the low 18F-FDG uptake. NAMC-CPET is advantageous in the low-statistic situation. The results are promising with great potential implications in cardiac PET imaging in terms of the radioactive dose and scan time reduction.

Accurate and robust reconstruction of the radioactivity concentration is of great importance in positron emission tomography (PET) imaging. Given the Poisson nature of photo-counting measurements, we present a reconstruction framework that integrates sparsity penalty on a dictionary into a maximum likelihood estimator. Patch-sparsity on a dictionary provides the regularization for our effort, and iterative procedures are used to solve the maximum likelihood function formulated on Poisson statistics. Specifically, in our formulation, a dictionary could be trained on CT images, to provide intrinsic anatomical structures for the reconstructed images, or adaptively learned from the noisy measurements of PET. Accuracy of the strategy with very promising application results from Monte-Carlo simulations, and real data are demonstrated.

The aim of this study is to investigate the impact of respiratory motion correction and spatial resolution on lesion detectability in PET as a function of lesion size and tracer uptake. Real respiratory signals describing different breathing types are combined with a motion model formed from real dynamic MR data to simulate multiple dynamic PET datasets acquired from a continuously moving subject. Lung and liver lesions were simulated with diameters ranging from 6 to 12 mm and lesion to background ratio ranging from 3:1 to 6:1. Projection data for 6 and 3 mm PET scanner resolution were generated using analytic simulations and reconstructed without and with motion correction. Motion correction was achieved using motion compensated image reconstruction. The detectability performance was quantified by a receiver operating characteristic (ROC) analysis obtained using a channelized Hotelling observer and the area under the ROC curve (AUC) was calculated as the figure of merit. The results indicate that respiratory motion limits the detectability of lung and liver lesions, depending on the variation of the breathing cycle length and amplitude. Patients with large quiescent periods had a greater AUC than patients with regular breathing cycles and patients with long-term variability in respiratory cycle or higher motion amplitude. In addition, small (less than 10 mm diameter) or low contrast (3:1) lesions showed the greatest improvement in AUC as a result of applying motion correction. In particular, after applying motion correction the AUC is improved by up to 42% with current PET resolution (i.e. 6 mm) and up to 51% for higher PET resolution (i.e. 3 mm). Finally, the benefit of increasing the scanner resolution is small unless motion correction is applied. This investigation indicates high impact of respiratory motion correction on lesion detectability in PET and highlights the importance of motion correction in order to benefit from the increased resolution of future PET scanners.

We investigated the severity of partial volume effects (PVE), which may occur in SPECT/CT studies, and the performance of two template-based correction techniques. A hybrid SPECT/CT system was used to scan a thorax phantom that included lungs, a heart insert and six cylindrical containers of different sizes and activity concentrations. This phantom configuration allowed us to have non-uniform background activity and a combination of spill-in and spill-out effects for several compartments. The reconstruction with corrections for attenuation, scatter and resolution loss but not PVE correction accurately recovered absolute activities in large organs. However, the activities inside segmented 17-120 mL containers were underestimated by 20%-40%. After applying our PVE correction to the data pertaining to six small containers, the accuracy of the recovered total activity improved with errors ranging between 3% and 22% (non-iterative method) and between 5% and 15% (method with an iteratively updated background activity). While the non-iterative template-based algorithm demonstrated slightly better accuracy for cases with less severe PVE than the iterative algorithm, it underperformed in situations with considerable spill out and/or mixture of spill-in and spill-out effects.

Respiratory motion results in significant motion blur in thoracic positron emission tomography (PET) imaging. Existing approaches to correct the blurring artifact involve acquiring the images in gated mode and using complicated reconstruction algorithms. In this paper, we propose a post-reconstruction framework to estimate respiratory motion and reduce the motion blur of PET images acquired in ungated mode. Our method includes two steps: one is to use minmax directional derivative analysis and local auto-correlation analysis to identify the two parameters blur direction and blur extent, respectively, and another is to employ WRL, à trous wavelet-denoising modified Richardson-Lucy (RL) deconvolution, to reduce the motion blur based on identified parameters. The mobile phantom data were first used to test the method before it was applied to 32 cases of clinical lung tumor PET data. Results showed that the blur extent of phantom images in different directions was accurately identified, and WRL can remove the majority of motion blur within ten iterations. The blur extent of clinical images was estimated to be 12.1 ± 3.7 mm in the direction of 74 ± 3° relative to the image horizontal axis. The quality of clinical images was significantly improved, both from visual inspection and quantitative evaluation after deconvolution. It was demonstrated that WRL outperforms RL and a Wiener filter in reducing the motion blur with one to two more iterations. The proposed method is easy to implement and thus could be a useful tool to reduce the effect of respiration in ungated thoracic PET imaging.

On a positron emission tomography (PET) scanner consisting of block detectors, coincidence responses to scattered radiation may differ from those to true depending on the crystal pair position within a coincidence block pair. Furthermore, these differences are considered to vary according to the radial position of the coincidence block pair. These conditions create ringing artifacts in the reconstructed image due to the lack of scatter compensation in detector normalization. In component-based normalization, a scatter-compensated crystal interference factor is therefore required in addition to the scatter-compensated block profile and intrinsic crystal efficiencies. In this study, we propose a scatter-compensated component-based normalization scheme using an annulus phantom, which provides true and scattered radiations over a large transaxial field of view, and evaluates the quality of three different-sized phantom images with whole-body PET. The results showed that the proposed normalization method significantly reduces the ringing artifacts in reconstructed images with different scattered/true fractions. The proposed algorithm, which introduced the scatter-compensated crystal interference factor, worked well under different scattered/true ratio conditions and was considered to be a robust, practical normalization method in high-resolution whole-body PET.

The incorporation of accurately aligned anatomical information as a prior to guide reconstruction and noise regularization in positron emission tomography (PET) has been suggested in many previous studies. However, the advantages of this approach can only be realized if the exact lesion outline is also available. In practice, the anatomical imaging modality may be unable to differentiate between normal and pathological tissues, and thus the edges of lesions seen in the anatomical image may not correspond to functional boundaries in the emission image. In this study, we explored an alternative approach to incorporating an anatomical prior into PET image reconstruction. Of particular interest was the realistic situation where lesions are apparent in the emission images but not in the corresponding anatomical images. In the proposed method, regional information obtained from the anatomical prior was used to estimate an anatomically adaptive anisotropic median-diffusion filtering (AAMDF) prior. This smoothing prior was determined and applied adaptively to each anatomical region on the emission image and then assembled to form a prior image for the next iteration in the reconstruction process. We formulated a two-step joint estimation reconstruction scheme to update the estimated image and prior image iteratively. The proposed AAMDF prior was evaluated and compared with maximum a posteriori (MAP) reconstruction methods with and without anatomical side information. In experiments using synthetic and physical phantom data, the AAMDF prior yielded overall higher lesion-to-background contrast and less error in lesion estimation than other algorithms for a comparable level of background noise. We conclude that lesion contrast and quantification can be improved using an anatomically derived smoothing prior without requiring knowledge of the lesion boundary. This may have important implications in clinical PET/CT, where lesion boundaries are often not obtainable from CT images.

Unlabelled: It is common, even with new SPECT/CT systems, that the transmission data are truncated. This paper develops a method that obtains exact attenuation correction with truncated transmission data. The emission object (e.g., the heart) is assumed to have a finite, convex support, whose emission projections are not truncated. The transmission measurements over the support are available, but may be truncated outside the support (within the torso). A novel emission data reconstruction technique combines emission projections from conjugate views; a modified version of the ML-EM algorithm is used to reconstruct emission data. The attenuation map outside the support is not needed during reconstruction. The transmission measurements through the support are used to pre-scale the emission data and to reconstruct the attenuation map within the support. The attenuation map reconstruction within the support is an interior problem in which only a biased solution can be obtained using an iterative algorithm. The bias is then corrected by identifying a soft tissue region within the support and the known attenuation coefficient values of these pixels for the soft tissue. Proof of convergence of the new algorithm is provided. Computer simulations verify the accuracy of the new method., Conclusions: an exact attenuation map within the support can be obtained provided the attenuation coefficient is known at 1 pixel within the support. The method, which requires emission data over 360 degrees , provides a means to perform attenuation correction in SPECT with truncated transmission data.

Partial volume effects (PVEs) are consequences of the limited spatial resolution in emission tomography. They lead to a loss of signal in tissues of size similar to the point spread function and induce activity spillover between regions. Although PVE can be corrected for by using algorithms that provide the correct radioactivity concentration in a series of regions of interest (ROIs), so far little attention has been given to the possibility of creating improved images as a result of PVE correction. Potential advantages of PVE-corrected images include the ability to accurately delineate functional volumes as well as improving tumour-to-background ratio, resulting in an associated improvement in the analysis of response to therapy studies and diagnostic examinations, respectively. The objective of our study was therefore to develop a methodology for PVE correction not only to enable the accurate recuperation of activity concentrations, but also to generate PVE-corrected images. In the multiresolution analysis that we define here, details of a high-resolution image H (MRI or CT) are extracted, transformed and integrated in a low-resolution image L (PET or SPECT). A discrete wavelet transform of both H and L images is performed by using the "à trous" algorithm, which allows the spatial frequencies (details, edges, textures) to be obtained easily at a level of resolution common to H and L. A model is then inferred to build the lacking details of L from the high-frequency details in H. The process was successfully tested on synthetic and simulated data, proving the ability to obtain accurately corrected images. Quantitative PVE correction was found to be comparable with a method considered as a reference but limited to ROI analyses. Visual improvement and quantitative correction were also obtained in two examples of clinical images, the first using a combined PET/CT scanner with a lymphoma patient and the second using a FDG brain PET and corresponding T1-weighted MRI in an epileptic patient.

Accurate estimation of the 3D in vivo activity distribution is important for dose estimation in targeted radionuclide therapy (TRT). Although SPECT can potentially provide such estimates, SPECT without compensation for image degrading factors is not quantitatively accurate. In this work, we evaluated quantitative SPECT (QSPECT) reconstruction methods that include compensation for various physical effects. Experimental projection data were obtained using a GE VH/Hawkeye system and an RSD torso phantom. Known activities of In-111 chloride were placed in the lungs, liver, heart, background and two spherical compartments with inner diameters of 22 mm and 34 mm. The 3D NCAT phantom with organ activities based on clinically derived In-111 ibritumomab tiuxetan data was used for the Monte Carlo (MC) simulation studies. Low-noise projection data were simulated using previously validated MC simulation methods. Fifty sets of noisy projections with realistic count levels were generated. Reconstructions were performed using the OS-EM algorithm with various combinations of attenuation (A), scatter (S), geometric response (G), collimator-detector response (D) and partial volume compensation (PVC). The QSPECT images from the various combinations of compensations were evaluated in terms of the accuracy and precision of the estimates of the total activity in each organ. For experimental data, the errors in organ activities for ADS and PVC compensation were less than 6.5% except the smaller sphere (-11.9%). For the noisy simulated data, the errors in organ activity for ADS compensation were less than 5.5% except the lungs (20.9%) and blood vessels (15.2%). Errors for other combinations of compensations were significantly (A, AS) or somewhat (AGS) larger. With added PVC, the error in the organ activities improved slightly except for the lungs (11.5%) and blood vessels (3.6%) where the improvement was more substantial. The standard deviation/mean ratios were all less than 1.5%. We conclude that QSPECT methods with appropriate compensations provided accurate In-111 organ activity estimates. For the collimator used, AGS was almost as good as ADS and may be preferable due to the reduced reconstruction time. PVC was important for small structures such as tumours or for organs in close proximity to regions with high activity. The improved quantitative accuracy from QSPECT methods has the potential for improving organ dose estimations in TRT.

We compare two different implementations of a 3D SPECT system model for iterative reconstruction, both of which compensate for non-uniform photon attenuation and depth-dependent system response. One implementation performs fast rotation of images represented using a basis of rectangular voxels, whereas the other represents images using a basis of rotationally symmetric volume elements. In our simulations the blob-based approach was found to slightly outperform the rotation-based one in terms of the bias-variance tradeoff in the reconstructed images. Their difference can be significant, however, in terms of computational load. The rotation-based method is faster for many typical SPECT reconstruction problems, but the blob-based one can be better-suited to cases where the reconstruction algorithm needs to process one volume element at a time.

The quality of thorax and pelvis transmission scans and therefore of attenuation correction in PET depends on patient thickness and transmission rod source strength. The purpose of the present study was to assess the feasibility of using count-based transmission scans, thereby guaranteeing more consistent image quality and more precise quantification than with fixed transmission scan duration. First, the relation between noise equivalent counts (NEC) of 10 min calibration transmission scans and rod source activity was determined over a period of 1.5 years. Second, the relation between transmission scan counts and uniform phantom diameter was studied numerically, determining the relative contribution of counts from lines of response passing through the phantom as compared with the total number of counts. Finally, the relation between patient weight and transmission scan duration was determined for 35 patients, who were scanned at the level of thorax or pelvis. After installation of new rod sources, the NEC of transmission scans first increased slightly (5%) with decreasing rod source activity and after 3 months decreased with a rate of 2-3% per month. The numerical simulation showed that the number of transmission scan counts from lines of response passing through the phantom increased with phantom diameter up to 7 cm. For phantoms larger than 7 cm, the number of these counts decreased at approximately the same rate as the total number of transmission scan counts. Patient data confirmed that the total number of transmission scan counts decreased with increasing patient weight with about 0.5% kg(-1). It can be concluded that count-based transmission scans compensate for radioactive decay of the rod sources. With count-based transmission scans, rod sources can be used for up to 1.5 years at the cost of a 50% increased transmission scan duration. For phantoms with diameters of more than 7 cm and for patients scanned at the level of thorax or pelvis, use of count-based transmission scans is feasible and results in statistically more consistent transmission scans as compared with fixed transmission scan duration.

The aim of this work is to assess the accuracy of a surface matching registration (SMR) technique for the correlation of cardiac studies in positron emission tomography (PET). Registration parameters were estimated by matching corresponding body surfaces, extracted from transmission studies, aligned to the PET emission images to be correlated. The accuracy of the SMR technique in this specific application was assessed by computer simulations, phantom experiments and on clinical PET data. Registration accuracy was evaluated in relation to the body surfaces (external, internal and the combination of the two) used by the SMR method. Better results were found when matching shaped and irregular surfaces such as internal lung contours. The robustness of the method was verified for different counting statistics recorded in transmission images. A clinical validation of the SMR method was performed on fluorine-18-deoxyglucose PET cardiac studies.