Software for Tomographic Image Reconstruction (STIR) is an open source C++ library used to recons... more Software for Tomographic Image Reconstruction (STIR) is an open source C++ library used to reconstruct single photon emission tomography and positron emission tomography (PET) data. STIR has an experimental scanner geometry modelling feature to accurately model detector placement. In this study, we test and improve this new feature using several types of data: Monte Carlo simulations and measured phantom data acquired from a dedicated brain PET prototype scanner. The results show that the new geometry class applied to non-cylindrical PET scanners improved spatial resolution, uniformity, and image contrast. These are directly observed in the reconstructions of small features in the test quality phantom. Overall, we conclude that the revised “BlocksOnCylindrical” class will be a valuable addition to the next STIR software release with adjustments of existing features (Single Scatter Simulation, forward projection, attenuation corrections) to “BlocksOnCylindrical”.
Synthesising computed tomography (CT) images from magnetic resonance images (MRI) plays an import... more Synthesising computed tomography (CT) images from magnetic resonance images (MRI) plays an important role in the field of medical image analysis, both for quantification and diagnostic purposes. Convolutional neural networks (CNNs) have achieved state-of-the-art results in image-to-image translation for brain applications. However, synthesising whole-body images remains largely uncharted territory, involving many challenges, including large image size and limited field of view, complex spatial context, and anatomical differences between images acquired at different times. We propose the use of an uncertainty-aware multi-channel multi-resolution 3D cascade network specifically aiming for whole-body MR to CT synthesis. The Mean Absolute Error on the synthetic CT generated with the MultiResunc network (73.90 HU) is compared to multiple baseline CNNs like 3D U-Net (92.89 HU), HighRes3DNet (89.05 HU) and deep boosted regression (77.58 HU) and shows superior synthesis performance. We ulti...
2017 IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC), 2017
This manuscript gives an update on the integration of Time-Of-Flight reconstruction into the STIR... more This manuscript gives an update on the integration of Time-Of-Flight reconstruction into the STIR image reconstruction toolkit. In this iteration we provide significant support to reconstruct TOF-compatible projection data. Most infrastructure classes, such as ProjDataInfo and ProjData, have been extended, and utilities, as lm_to_projdata, used to create sinograms from listmode data, can now handle TOF information. This extension required many modifications in the low level code base of STIR, making it non-trivial and error-prone. Therefore, a thorough validation is required. In this work we provide initial results of the correctness of the extension. Using Monte Carlo simulations, as well as cylindrical and XCAT phantoms analytically projected, we calculate the contrast recovery ratio over a wide range of iterations. The results demonstrate the benefits of TOF under different configurations, which are in good agreement with literature. The ROI mean value converges to the maximum fa...
Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2021
SIRF is a powerful PET/MR image reconstruction research tool for processing data and developing n... more SIRF is a powerful PET/MR image reconstruction research tool for processing data and developing new algorithms. In this research, new developments to SIRF are presented, with focus on motion estimation and correction. SIRF’s recent inclusion of the adjoint of the resampling operator allows gradient propagation through resampling, enabling the MCIR technique. Another enhancement enabled registering and resampling of complex images, suitable for MRI. Furthermore, SIRF’s integration with the optimization library CIL enables the use of novel algorithms. Finally, SPM is now supported, in addition to NiftyReg, for registration. Results of MR and PET MCIR reconstructions are presented, using FISTA and PDHG, respectively. These demonstrate the advantages of incorporating motion correction and variational and structural priors. This article is part of the theme issue ‘Synergistic tomographic image reconstruction: part 2’.
Kinetic analysis of dynamic PET data requires an accurate knowledge of available PET tracer conce... more Kinetic analysis of dynamic PET data requires an accurate knowledge of available PET tracer concentration within blood plasma over time, known as the arterial input function (AIF). The gold standard method used to measure the AIF requires serial arterial blood sampling over the course of the PET scan, which is an invasive procedure and makes this method less practical in clinical settings. Traditional image-derived methods are limited to specific tracers and are not accurate if metabolites are present in the plasma. In this work, we utilise an image-derived whole blood curve measurement to reduce the computational complexity of the simultaneous estimation method (SIME), which is capable of estimating the AIF directly from tissue time activity curves (TACs). This method was applied to data obtained from a serotonin receptor study (C-SB207145) and estimated parameter results are compared to results obtained using the original SIME and gold standard AIFs derived from arterial samples. ...
Journal of nuclear medicine : official publication, Society of Nuclear Medicine, Jan 9, 2018
In Positron Emission Tomography (PET) imaging, patient motion due to respiration can lead to arte... more In Positron Emission Tomography (PET) imaging, patient motion due to respiration can lead to artefacts and blurring, in addition to quantification errors. The integration of PET imaging with Magnetic Resonance (MR) imaging in PET/MR scanners provides spatially aligned complementary clinical information, and allows the use of high spatial resolution and high contrast MR images to monitor and correct motion-corrupted PET data. We validate our PET respiratory motion correction methodology based on a joint PET-MR motion model, on a patient cohort, showing it can improve lesion detectability and quantitation, and reduce image artefacts.We apply our motion correction methodology on 42 clinical PET-MR patient datasets, using multiple tracers and multiple organ locations, containing 162 PET-avid lesions. Quantitative changes are calculated using Standardised Uptake Value (SUV) changes in avid lesions. Lesion detectability changes are explored with a study where two radiologists identify les...
European journal of nuclear medicine and molecular imaging, Jan 16, 2018
There is a lack of prognostic biomarkers in idiopathic pulmonary fibrosis (IPF) patients. The obj... more There is a lack of prognostic biomarkers in idiopathic pulmonary fibrosis (IPF) patients. The objective of this study is to investigate the potential of 18F-FDG-PET/ CT to predict mortality in IPF. A total of 113 IPF patients (93 males, 20 females, mean age ± SD: 70 ± 9 years) were prospectively recruited for 18F-FDG-PET/CT. The overall maximum pulmonary uptake of 18F-FDG (SUVmax), the minimum pulmonary uptake or background lung activity (SUVmin), and target-to-background (SUVmax/ SUVmin) ratio (TBR) were quantified using routine region-of-interest analysis. Kaplan-Meier analysis was used to identify associations of PET measurements with mortality. We also compared PET associations with IPF mortality with the established GAP (gender age and physiology) scoring system. Cox analysis assessed the independence of the significant PET measurement(s) from GAP score. We investigated synergisms between pulmonary 18F-FDG-PET measurements and GAP score for risk stratification in IPF patients. ...
Respiratory motion compensation in PET/CT and PET/MRI is essential as motion is a source of image... more Respiratory motion compensation in PET/CT and PET/MRI is essential as motion is a source of image degradation (motion blur, attenuation artifacts). In previous work, we developed a direct method for joint image reconstruction/motion estimation (JRM) for attenuation-corrected (AC) respiratory-gated PET, which uses a single attenuation-map (μ-map). This approach was successfully implemented for respiratory-gated PET/CT, but since it relied on an accurate μ-map for motion estimation, the question of its applicability in PET/MRI is open. The purpose of this work is to investigate the feasibility of JRM in PET/MRI and to assess the robustness of the motion estimation when a degraded μ-map is used. We performed a series of JRM reconstructions from simulated PET data using a range of simulated Dixon MRI sequence derived μ-maps with wrong attenuation values in the lungs, from - 100% (no attenuation) to +100% (double attenuation), as well as truncated arms. We compared the estimated motions ...
Journal of nuclear medicine : official publication, Society of Nuclear Medicine, Feb 1, 2017
Millions of people are affected by respiratory diseases, leading to a significant health burden g... more Millions of people are affected by respiratory diseases, leading to a significant health burden globally. Because of the current insufficient knowledge of the underlying mechanisms that lead to the development and progression of respiratory diseases, treatment options remain limited. To overcome this limitation and understand the associated molecular changes, noninvasive imaging techniques such as PET and SPECT have been explored for biomarker development, with (18)F-FDG PET imaging being the most studied. The quantification of pulmonary molecular imaging data remains challenging because of variations in tissue, air, blood, and water fractions within the lungs. The proportions of these components further differ depending on the lung disease. Therefore, different quantification approaches have been proposed to address these variabilities. However, no standardized approach has been developed to date. This article reviews the data evaluating (18)F-FDG PET quantification approaches in l...
Software for Tomographic Image Reconstruction (STIR) is an open source C++ library used to recons... more Software for Tomographic Image Reconstruction (STIR) is an open source C++ library used to reconstruct single photon emission tomography and positron emission tomography (PET) data. STIR has an experimental scanner geometry modelling feature to accurately model detector placement. In this study, we test and improve this new feature using several types of data: Monte Carlo simulations and measured phantom data acquired from a dedicated brain PET prototype scanner. The results show that the new geometry class applied to non-cylindrical PET scanners improved spatial resolution, uniformity, and image contrast. These are directly observed in the reconstructions of small features in the test quality phantom. Overall, we conclude that the revised “BlocksOnCylindrical” class will be a valuable addition to the next STIR software release with adjustments of existing features (Single Scatter Simulation, forward projection, attenuation corrections) to “BlocksOnCylindrical”.
Synthesising computed tomography (CT) images from magnetic resonance images (MRI) plays an import... more Synthesising computed tomography (CT) images from magnetic resonance images (MRI) plays an important role in the field of medical image analysis, both for quantification and diagnostic purposes. Convolutional neural networks (CNNs) have achieved state-of-the-art results in image-to-image translation for brain applications. However, synthesising whole-body images remains largely uncharted territory, involving many challenges, including large image size and limited field of view, complex spatial context, and anatomical differences between images acquired at different times. We propose the use of an uncertainty-aware multi-channel multi-resolution 3D cascade network specifically aiming for whole-body MR to CT synthesis. The Mean Absolute Error on the synthetic CT generated with the MultiResunc network (73.90 HU) is compared to multiple baseline CNNs like 3D U-Net (92.89 HU), HighRes3DNet (89.05 HU) and deep boosted regression (77.58 HU) and shows superior synthesis performance. We ulti...
2017 IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC), 2017
This manuscript gives an update on the integration of Time-Of-Flight reconstruction into the STIR... more This manuscript gives an update on the integration of Time-Of-Flight reconstruction into the STIR image reconstruction toolkit. In this iteration we provide significant support to reconstruct TOF-compatible projection data. Most infrastructure classes, such as ProjDataInfo and ProjData, have been extended, and utilities, as lm_to_projdata, used to create sinograms from listmode data, can now handle TOF information. This extension required many modifications in the low level code base of STIR, making it non-trivial and error-prone. Therefore, a thorough validation is required. In this work we provide initial results of the correctness of the extension. Using Monte Carlo simulations, as well as cylindrical and XCAT phantoms analytically projected, we calculate the contrast recovery ratio over a wide range of iterations. The results demonstrate the benefits of TOF under different configurations, which are in good agreement with literature. The ROI mean value converges to the maximum fa...
Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2021
SIRF is a powerful PET/MR image reconstruction research tool for processing data and developing n... more SIRF is a powerful PET/MR image reconstruction research tool for processing data and developing new algorithms. In this research, new developments to SIRF are presented, with focus on motion estimation and correction. SIRF’s recent inclusion of the adjoint of the resampling operator allows gradient propagation through resampling, enabling the MCIR technique. Another enhancement enabled registering and resampling of complex images, suitable for MRI. Furthermore, SIRF’s integration with the optimization library CIL enables the use of novel algorithms. Finally, SPM is now supported, in addition to NiftyReg, for registration. Results of MR and PET MCIR reconstructions are presented, using FISTA and PDHG, respectively. These demonstrate the advantages of incorporating motion correction and variational and structural priors. This article is part of the theme issue ‘Synergistic tomographic image reconstruction: part 2’.
Kinetic analysis of dynamic PET data requires an accurate knowledge of available PET tracer conce... more Kinetic analysis of dynamic PET data requires an accurate knowledge of available PET tracer concentration within blood plasma over time, known as the arterial input function (AIF). The gold standard method used to measure the AIF requires serial arterial blood sampling over the course of the PET scan, which is an invasive procedure and makes this method less practical in clinical settings. Traditional image-derived methods are limited to specific tracers and are not accurate if metabolites are present in the plasma. In this work, we utilise an image-derived whole blood curve measurement to reduce the computational complexity of the simultaneous estimation method (SIME), which is capable of estimating the AIF directly from tissue time activity curves (TACs). This method was applied to data obtained from a serotonin receptor study (C-SB207145) and estimated parameter results are compared to results obtained using the original SIME and gold standard AIFs derived from arterial samples. ...
Journal of nuclear medicine : official publication, Society of Nuclear Medicine, Jan 9, 2018
In Positron Emission Tomography (PET) imaging, patient motion due to respiration can lead to arte... more In Positron Emission Tomography (PET) imaging, patient motion due to respiration can lead to artefacts and blurring, in addition to quantification errors. The integration of PET imaging with Magnetic Resonance (MR) imaging in PET/MR scanners provides spatially aligned complementary clinical information, and allows the use of high spatial resolution and high contrast MR images to monitor and correct motion-corrupted PET data. We validate our PET respiratory motion correction methodology based on a joint PET-MR motion model, on a patient cohort, showing it can improve lesion detectability and quantitation, and reduce image artefacts.We apply our motion correction methodology on 42 clinical PET-MR patient datasets, using multiple tracers and multiple organ locations, containing 162 PET-avid lesions. Quantitative changes are calculated using Standardised Uptake Value (SUV) changes in avid lesions. Lesion detectability changes are explored with a study where two radiologists identify les...
European journal of nuclear medicine and molecular imaging, Jan 16, 2018
There is a lack of prognostic biomarkers in idiopathic pulmonary fibrosis (IPF) patients. The obj... more There is a lack of prognostic biomarkers in idiopathic pulmonary fibrosis (IPF) patients. The objective of this study is to investigate the potential of 18F-FDG-PET/ CT to predict mortality in IPF. A total of 113 IPF patients (93 males, 20 females, mean age ± SD: 70 ± 9 years) were prospectively recruited for 18F-FDG-PET/CT. The overall maximum pulmonary uptake of 18F-FDG (SUVmax), the minimum pulmonary uptake or background lung activity (SUVmin), and target-to-background (SUVmax/ SUVmin) ratio (TBR) were quantified using routine region-of-interest analysis. Kaplan-Meier analysis was used to identify associations of PET measurements with mortality. We also compared PET associations with IPF mortality with the established GAP (gender age and physiology) scoring system. Cox analysis assessed the independence of the significant PET measurement(s) from GAP score. We investigated synergisms between pulmonary 18F-FDG-PET measurements and GAP score for risk stratification in IPF patients. ...
Respiratory motion compensation in PET/CT and PET/MRI is essential as motion is a source of image... more Respiratory motion compensation in PET/CT and PET/MRI is essential as motion is a source of image degradation (motion blur, attenuation artifacts). In previous work, we developed a direct method for joint image reconstruction/motion estimation (JRM) for attenuation-corrected (AC) respiratory-gated PET, which uses a single attenuation-map (μ-map). This approach was successfully implemented for respiratory-gated PET/CT, but since it relied on an accurate μ-map for motion estimation, the question of its applicability in PET/MRI is open. The purpose of this work is to investigate the feasibility of JRM in PET/MRI and to assess the robustness of the motion estimation when a degraded μ-map is used. We performed a series of JRM reconstructions from simulated PET data using a range of simulated Dixon MRI sequence derived μ-maps with wrong attenuation values in the lungs, from - 100% (no attenuation) to +100% (double attenuation), as well as truncated arms. We compared the estimated motions ...
Journal of nuclear medicine : official publication, Society of Nuclear Medicine, Feb 1, 2017
Millions of people are affected by respiratory diseases, leading to a significant health burden g... more Millions of people are affected by respiratory diseases, leading to a significant health burden globally. Because of the current insufficient knowledge of the underlying mechanisms that lead to the development and progression of respiratory diseases, treatment options remain limited. To overcome this limitation and understand the associated molecular changes, noninvasive imaging techniques such as PET and SPECT have been explored for biomarker development, with (18)F-FDG PET imaging being the most studied. The quantification of pulmonary molecular imaging data remains challenging because of variations in tissue, air, blood, and water fractions within the lungs. The proportions of these components further differ depending on the lung disease. Therefore, different quantification approaches have been proposed to address these variabilities. However, no standardized approach has been developed to date. This article reviews the data evaluating (18)F-FDG PET quantification approaches in l...
Uploads
Papers by Kris Thielemans