Neutron imaging is complementary to x-ray imaging in that materials such as water and plastic are... more Neutron imaging is complementary to x-ray imaging in that materials such as water and plastic are highly attenuating while material such as metal is nearly transparent. We showcase tomographic imaging of a diesel particulate filter. Reconstruction is done using a modified version of SIRT called PSIRT. We expand on previous work and introduce Tikhonov regularization. We show that near-optimal relaxation can still be achieved. The algorithmic ideas apply to cone beam x-ray CT and other inverse problems.
Positron Emission Tomography (PET) is a diagnostic technique used to study functionality of human... more Positron Emission Tomography (PET) is a diagnostic technique used to study functionality of human organs. A maximum likelihood reconstruction algorithm for PET, introduced by Shepp and Vardi and Lange and Carson, produces images that are statistically superior to other methods, but is an iterative method with high time complexity. Most efforts at parallelization have been for single instruction stream, multiple data stream (SIMD) parallel computers. This research extends the mathematical development of the iterative reconstruction and presents new results in parallel multiple instruction stream, multiple data stream (MIMD) computation with reduced processor communications. We obtain alternative convergence results and new results for rate of convergence of the iterative reconstruction algorithm. We show that the algorithm can be parallelized with reduced communications and show empirically using both simulated and clinical data that the reduced communication algorithm produces reconstructions that are comparable to reconstructions computed with full communication.
2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2018
Respiratory motion during PET/CT imaging is a matter of concern due to degraded image quality and... more Respiratory motion during PET/CT imaging is a matter of concern due to degraded image quality and reduced quantitative accuracy caused by motion artifacts. One class of motion correction methods relies on hardware-based respiratory motion tracking systems in order to use respiratory cycles for correcting motion artifacts. Another class of hardware-free methods extract motion information from the reconstructed images or sinograms. Hardware-based methods, however, are limited by calibration requirement, patient discomfort, lack of adaptability during scanning, presence of electronic drift during respiratory monitoring etc. Extracting motion information from reconstructed images is also limited by the fact that the original raw information requires significant processing before it can be used. Hence the motivation behind this work is to introduce a software-based approach that can be applied on raw 64-bit listmode data. The basic design of the proposed method is based on the fundamentals of Positron Emission Particle Tracking (PEPT) with additional incorporation of Time of Flight (TOF) information. Respiratory motion of patients has been extracted from the raw PET data by tracking a point source attached to the patient in areas on and near the chest. The key objective of this work is to describe a new process by which this particle tracking based motion correction system can eventually be lesion specific and correct the motion for a particular lesion within the patient. This work thus serves as a framework for lesion specific motion correction.
The use of x-ray computed tomography (CT) scanners has become widespread in both clinical and pre... more The use of x-ray computed tomography (CT) scanners has become widespread in both clinical and preclinical contexts. CT scanners can be used to noninvasively test for anatomical anomalies as well as to diagnose and monitor disease progression. However, the data acquired by a ...
Purpose Respiratory motion of patients during positron emission tomography (PET)/computed tomogra... more Purpose Respiratory motion of patients during positron emission tomography (PET)/computed tomography (CT) imaging affects both image quality and quantitative accuracy. Hardware‐based motion estimation, which is the current clinical standard, requires initial setup, maintenance, and calibration of the equipment, and can be associated with patient discomfort. Data‐driven techniques are an active area of research with limited exploration into lesion‐specific motion estimation. This paper introduces a time‐of‐flight (TOF)‐weighted positron emission particle tracking (PEPT) algorithm that facilitates lesion‐specific respiratory motion estimation from raw listmode PET data. Methods The TOF‐PEPT algorithm was implemented and investigated under different scenarios: (a) a phantom study with a point source and an Anzai band for respiratory motion tracking; (b) a phantom study with a point source only, no Anzai band; (c) two clinical studies with point sources and the Anzai band; (d) two clinical studies with point sources only, no Anzai band; and (e) two clinical studies using lesions/internal regions instead of point sources and no Anzai band. For studies with radioactive point sources, they were placed on patients during PET/CT imaging. The motion tracking was performed using a preselected region of interest (ROI), manually drawn around point sources or lesions on reconstructed images. The extracted motion signals were compared with the Anzai band when applicable. For the purposes of additional comparison, a center‐of‐mass (COM) algorithm was implemented both with and without the use of TOF information. Using the motion estimate from each method, amplitude‐based gating was applied, and gated images were reconstructed. Results The TOF‐PEPT algorithm is shown to successfully determine the respiratory motion for both phantom and clinical studies. The derived motion signals correlated well with the Anzai band; correlation coefficients of 0.99 and 0.94‐0.97 were obtained for the phantom study and the clinical studies, respectively. TOF‐PEPT was found to be 13–38% better correlated with the Anzai results than the COM methods. Maximum Standardized Uptake Values (SUVs) were used to quantitatively compare the reconstructed‐gated images. In comparison with the ungated image, a 14–39% increase in the max SUV across several lesion areas and an 8.7% increase in the max SUV on the tracked lesion area were observed in the gated images based on TOF‐PEPT. The distinct presence of lesions with reduced blurring effect and generally sharper images were readily apparent in all clinical studies. In addition, max SUVs were found to be 4–10% higher in the TOF‐PEPT‐based gated images than in those based on Anzai and COM methods. Conclusion A PEPT‐ based algorithm has been presented for determining movement due to respiratory motion during PET/CT imaging. Gating based on the motion estimate is shown to quantifiably improve the image quality in both a controlled point source phantom study and in clinical data patient studies. The algorithm has the potential to facilitate true motion correction where the reconstruction algorithm can use all data available.
2019 IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC)
Direct neural network image reconstructions methods from raw data are notoriously memory ineffici... more Direct neural network image reconstructions methods from raw data are notoriously memory inefficient in the transformation from the measurement space domain to image space by use of one or more fully connected layers. This inefficiency has thus far limited the size of of the final image and posed a significant obstacle to studying direct neural network reconstruction methods. However, this memory space challenge can be corrected by observing the measurement data that contributes to the formation of a given image patch and designing a network layer to exploit those observations. This article takes this approach and proposes a novel Radon inversion neural network layer that enables full size direct image reconstruction from Radon encoded raw measurement data with a 99.84% decrease in memory requirements compared to the recently published AUTOMAP direct reconstruction network. We demonstrate high quality full size image reconstruction from a neural network using the Radon inversion layer and compare the results to the SART reconstruction method utilizing a structural similarity measure for comparison.
At the present, neutron sources cannot be fabricated small and powerful enough in order to achiev... more At the present, neutron sources cannot be fabricated small and powerful enough in order to achieve high resolution radiography while maintaining an adequate flux. One solution is to employ computational imaging techniques such as a Magnified Coded Source Imaging (CSI) system. A coded-mask is placed between the neutron source and the object. The system resolution is increased by reducing the size of the mask holes and the flux is increased by increasing the size of the coded-mask and/or the number of holes. One limitation of such system is that the resolution of current state-of-the-art scintillator-based detectors caps around 50um. To overcome this challenge, the coded-mask and object are magnified by making the distance from the coded-mask to the object much smaller than the distance from object to detector. In previous work, we have shown via synthetic experiments that our least squares method outperforms other methods in image quality and reconstruction precision because of the m...
IEEE Transactions on Radiation and Plasma Medical Sciences, 2021
Direct reconstruction of positron emission tomography (PET) data using deep neural networks is a ... more Direct reconstruction of positron emission tomography (PET) data using deep neural networks is a growing field of research. Initial results are promising, but often the networks are complex, memory utilization inefficient, produce relatively small 2-D image slices (e.g., $128 \times 128$ ), and low count rate reconstructions are of varying quality. This article proposes FastPET, a novel direct reconstruction convolutional neural network that is architecturally simple, memory space efficient, works for nontrivial 3-D image volumes and is capable of processing a wide spectrum of PET data including low-dose and multitracer applications. FastPET uniquely operates on a histo-image (i.e., image-space) representation of the raw data enabling it to reconstruct 3-D image volumes $67\times $ faster than ordered subsets expectation maximization (OSEM). We detail the FastPET method trained on whole-body and low-dose whole-body data sets and explore qualitative and quantitative aspects of reconstructed images from clinical and phantom studies. Additionally, we explore the application of FastPET on a neurology data set containing multiple different tracers. The results show that not only are the reconstructions very fast, but the images are high quality and have lower noise than iterative reconstructions.
Neutron imaging is complementary to x-ray imaging in that materials such as water and plastic are... more Neutron imaging is complementary to x-ray imaging in that materials such as water and plastic are highly attenuating while material such as metal is nearly transparent. We showcase tomographic imaging of a diesel particulate filter. Reconstruction is done using a modified version of SIRT called PSIRT. We expand on previous work and introduce Tikhonov regularization. We show that near-optimal relaxation can still be achieved. The algorithmic ideas apply to cone beam x-ray CT and other inverse problems.
Positron Emission Tomography (PET) is a diagnostic technique used to study functionality of human... more Positron Emission Tomography (PET) is a diagnostic technique used to study functionality of human organs. A maximum likelihood reconstruction algorithm for PET, introduced by Shepp and Vardi and Lange and Carson, produces images that are statistically superior to other methods, but is an iterative method with high time complexity. Most efforts at parallelization have been for single instruction stream, multiple data stream (SIMD) parallel computers. This research extends the mathematical development of the iterative reconstruction and presents new results in parallel multiple instruction stream, multiple data stream (MIMD) computation with reduced processor communications. We obtain alternative convergence results and new results for rate of convergence of the iterative reconstruction algorithm. We show that the algorithm can be parallelized with reduced communications and show empirically using both simulated and clinical data that the reduced communication algorithm produces reconstructions that are comparable to reconstructions computed with full communication.
2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2018
Respiratory motion during PET/CT imaging is a matter of concern due to degraded image quality and... more Respiratory motion during PET/CT imaging is a matter of concern due to degraded image quality and reduced quantitative accuracy caused by motion artifacts. One class of motion correction methods relies on hardware-based respiratory motion tracking systems in order to use respiratory cycles for correcting motion artifacts. Another class of hardware-free methods extract motion information from the reconstructed images or sinograms. Hardware-based methods, however, are limited by calibration requirement, patient discomfort, lack of adaptability during scanning, presence of electronic drift during respiratory monitoring etc. Extracting motion information from reconstructed images is also limited by the fact that the original raw information requires significant processing before it can be used. Hence the motivation behind this work is to introduce a software-based approach that can be applied on raw 64-bit listmode data. The basic design of the proposed method is based on the fundamentals of Positron Emission Particle Tracking (PEPT) with additional incorporation of Time of Flight (TOF) information. Respiratory motion of patients has been extracted from the raw PET data by tracking a point source attached to the patient in areas on and near the chest. The key objective of this work is to describe a new process by which this particle tracking based motion correction system can eventually be lesion specific and correct the motion for a particular lesion within the patient. This work thus serves as a framework for lesion specific motion correction.
The use of x-ray computed tomography (CT) scanners has become widespread in both clinical and pre... more The use of x-ray computed tomography (CT) scanners has become widespread in both clinical and preclinical contexts. CT scanners can be used to noninvasively test for anatomical anomalies as well as to diagnose and monitor disease progression. However, the data acquired by a ...
Purpose Respiratory motion of patients during positron emission tomography (PET)/computed tomogra... more Purpose Respiratory motion of patients during positron emission tomography (PET)/computed tomography (CT) imaging affects both image quality and quantitative accuracy. Hardware‐based motion estimation, which is the current clinical standard, requires initial setup, maintenance, and calibration of the equipment, and can be associated with patient discomfort. Data‐driven techniques are an active area of research with limited exploration into lesion‐specific motion estimation. This paper introduces a time‐of‐flight (TOF)‐weighted positron emission particle tracking (PEPT) algorithm that facilitates lesion‐specific respiratory motion estimation from raw listmode PET data. Methods The TOF‐PEPT algorithm was implemented and investigated under different scenarios: (a) a phantom study with a point source and an Anzai band for respiratory motion tracking; (b) a phantom study with a point source only, no Anzai band; (c) two clinical studies with point sources and the Anzai band; (d) two clinical studies with point sources only, no Anzai band; and (e) two clinical studies using lesions/internal regions instead of point sources and no Anzai band. For studies with radioactive point sources, they were placed on patients during PET/CT imaging. The motion tracking was performed using a preselected region of interest (ROI), manually drawn around point sources or lesions on reconstructed images. The extracted motion signals were compared with the Anzai band when applicable. For the purposes of additional comparison, a center‐of‐mass (COM) algorithm was implemented both with and without the use of TOF information. Using the motion estimate from each method, amplitude‐based gating was applied, and gated images were reconstructed. Results The TOF‐PEPT algorithm is shown to successfully determine the respiratory motion for both phantom and clinical studies. The derived motion signals correlated well with the Anzai band; correlation coefficients of 0.99 and 0.94‐0.97 were obtained for the phantom study and the clinical studies, respectively. TOF‐PEPT was found to be 13–38% better correlated with the Anzai results than the COM methods. Maximum Standardized Uptake Values (SUVs) were used to quantitatively compare the reconstructed‐gated images. In comparison with the ungated image, a 14–39% increase in the max SUV across several lesion areas and an 8.7% increase in the max SUV on the tracked lesion area were observed in the gated images based on TOF‐PEPT. The distinct presence of lesions with reduced blurring effect and generally sharper images were readily apparent in all clinical studies. In addition, max SUVs were found to be 4–10% higher in the TOF‐PEPT‐based gated images than in those based on Anzai and COM methods. Conclusion A PEPT‐ based algorithm has been presented for determining movement due to respiratory motion during PET/CT imaging. Gating based on the motion estimate is shown to quantifiably improve the image quality in both a controlled point source phantom study and in clinical data patient studies. The algorithm has the potential to facilitate true motion correction where the reconstruction algorithm can use all data available.
2019 IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC)
Direct neural network image reconstructions methods from raw data are notoriously memory ineffici... more Direct neural network image reconstructions methods from raw data are notoriously memory inefficient in the transformation from the measurement space domain to image space by use of one or more fully connected layers. This inefficiency has thus far limited the size of of the final image and posed a significant obstacle to studying direct neural network reconstruction methods. However, this memory space challenge can be corrected by observing the measurement data that contributes to the formation of a given image patch and designing a network layer to exploit those observations. This article takes this approach and proposes a novel Radon inversion neural network layer that enables full size direct image reconstruction from Radon encoded raw measurement data with a 99.84% decrease in memory requirements compared to the recently published AUTOMAP direct reconstruction network. We demonstrate high quality full size image reconstruction from a neural network using the Radon inversion layer and compare the results to the SART reconstruction method utilizing a structural similarity measure for comparison.
At the present, neutron sources cannot be fabricated small and powerful enough in order to achiev... more At the present, neutron sources cannot be fabricated small and powerful enough in order to achieve high resolution radiography while maintaining an adequate flux. One solution is to employ computational imaging techniques such as a Magnified Coded Source Imaging (CSI) system. A coded-mask is placed between the neutron source and the object. The system resolution is increased by reducing the size of the mask holes and the flux is increased by increasing the size of the coded-mask and/or the number of holes. One limitation of such system is that the resolution of current state-of-the-art scintillator-based detectors caps around 50um. To overcome this challenge, the coded-mask and object are magnified by making the distance from the coded-mask to the object much smaller than the distance from object to detector. In previous work, we have shown via synthetic experiments that our least squares method outperforms other methods in image quality and reconstruction precision because of the m...
IEEE Transactions on Radiation and Plasma Medical Sciences, 2021
Direct reconstruction of positron emission tomography (PET) data using deep neural networks is a ... more Direct reconstruction of positron emission tomography (PET) data using deep neural networks is a growing field of research. Initial results are promising, but often the networks are complex, memory utilization inefficient, produce relatively small 2-D image slices (e.g., $128 \times 128$ ), and low count rate reconstructions are of varying quality. This article proposes FastPET, a novel direct reconstruction convolutional neural network that is architecturally simple, memory space efficient, works for nontrivial 3-D image volumes and is capable of processing a wide spectrum of PET data including low-dose and multitracer applications. FastPET uniquely operates on a histo-image (i.e., image-space) representation of the raw data enabling it to reconstruct 3-D image volumes $67\times $ faster than ordered subsets expectation maximization (OSEM). We detail the FastPET method trained on whole-body and low-dose whole-body data sets and explore qualitative and quantitative aspects of reconstructed images from clinical and phantom studies. Additionally, we explore the application of FastPET on a neurology data set containing multiple different tracers. The results show that not only are the reconstructions very fast, but the images are high quality and have lower noise than iterative reconstructions.
Uploads
Papers by Jens Gregor