A5030318542- Medical Imaging Techniques and Applications
- Radiomics and Machine Learning in Medical Imaging
- Advanced X-ray and CT Imaging
- Advanced MRI Techniques and Applications
- Radiation Detection and Scintillator Technologies
- Radiopharmaceutical Chemistry and Applications
- Advanced Radiotherapy Techniques
- Prostate Cancer Treatment and Research
- Image Processing and 3D Reconstruction
- Advanced Semiconductor Detectors and Materials
- Particle Detector Development and Performance
- Prostate Cancer Diagnosis and Treatment
- Atomic and Subatomic Physics Research
- Artificial Intelligence in Healthcare and Education
- Terahertz technology and applications
- Cancer Genomics and Diagnostics
- MRI in cancer diagnosis
- Industrial Gas Emission Control
- Medical Image Segmentation Techniques
- Digital Radiography and Breast Imaging
- Cardiac Imaging and Diagnostics
- Lung Cancer Diagnosis and Treatment
- Oil, Gas, and Environmental Issues
- Chemical Looping and Thermochemical Processes
- Cancer Immunotherapy and Biomarkers
Siemens Healthcare (United States)
2019-2021
General Electric (Israel)
2021
General Electric (United States)
2006-2020
Stanford University
2019
University of Washington
2014-2018
GE Global Research (United States)
2008
We have previously developed a convergent penalized likelihood (PL) image reconstruction algorithm using the relative difference prior (RDP) and showed that it achieves more accurate lesion quantitation compared to ordered subsets expectation maximization (OSEM). evaluated detectability of low-contrast liver lung lesions PL-RDP OSEM. performed two-alternative forced choice study channelized Hotelling observer model was validated against human observers. Lesion stronger dependence on size for...
Gallium-68-labeled radiopharmaceuticals pose a challenge for scatter estimation because their targeted nature can produce high contrast in these regions of the kidneys and bladder. Even small errors estimate result washout artifacts. Administration diuretics reduce artifacts, but they may adverse events. Here, we investigated ability algorithmic modifications to mitigate artifacts eliminate need or other interventions.
Uptake time (interval between tracer injection and image acquisition) affects the SUV measured for tumors in <sup>18</sup>F-FDG PET images. With dissimilar uptake times, changes tumor SUVs will be under- or overestimated. This study examined influence of on response assessment using a virtual clinical trials approach. <b>Methods:</b> Tumor kinetic parameters were estimated from dynamic scans breast cancer patients used to simulate time–activity curves 45–120 min after injection. Five-minute...
This work demonstrates how computational and physical modelling of the positron emission tomography (PET) image acquisition process for a state-of-the-art integrated PET magnetic resonance imaging (PET-MR) system can produce images comparable to manufacturer. The GE SIGNA PET/MR scanner is manufactured by General Electric has time-of-flight (TOF) capabilities about 390 ps. All software development took place in Software Tomographic Image Reconstruction (STIR: http://stir.sf.net) library,...
We developed a method to evaluate variations in the PET imaging process order characterize relative ability of static and dynamic metrics measure breast cancer response therapy clinical trial setting. performed virtual by generating 540 independent identically distributed study realizations for each 22 original fluorodeoxyglucose (18F-FDG) patient studies pre- post-therapy. Each noise realization accounted known sources uncertainty process, such as biological variability SUV uptake time....
Historically, patient datasets have been used to develop and validate various reconstruction algorithms for PET/MRI PET/CT. To enable such algorithm development, without the need acquiring hundreds of exams, in this article we demonstrate a deep learning technique generate synthetic but realistic whole-body PET sinograms from abundantly available MRI. Specifically, use dataset 56 18F-FDG-PET/MRI exams train 3-D residual UNet predict physiologic uptake T1-weighted In training, implemented...
Prior reports have suggested that delayed FDG-PET oncology imaging can improve the contrast-to-noise ratio (CNR) for known lesions. Our goal was to estimate realistic bounds lesion detectability static measurements with one four hours between FDG injection and image acquisition. Tumor normal tissue kinetic model parameters were estimated from dynamic PET studies of patients early stage breast cancer. These used generate time-activity curves (TACs) out hours, which we assumed both...
Abstract Background In modern positron emission tomography (PET) with multi‐modality imaging (e.g., PET/CT and PET/MR), the attenuation correction (AC) is single largest factor for image reconstruction. One way to assess AC methods other reconstruction parameters utilize software‐based simulation tools, such as a lesion insertion tool. Extensive validation of these tools required ensure results study are clinically meaningful. Purpose To evaluate different PET using synthetic tool that...
An electrical model is developed to simulate, characterize, and predict the response of SSPM detectors for different device geometries measurement circuit configurations. In particular, allows investigation effects increasing parasitic capacitance with diode area on timing magnitude readout signal. Passive components in are extracted from measurements then used understand performance. The avalanche represented a switch series voltage source resistor, instead current source. This approach...
Ordered Subset Expectation Maximization (OSEM) is currently the most widely used image reconstruction algorithm for clinical PET. However, OSEM does not necessarily provide optimal quality, and a number of alternative algorithms have been explored. We recently shown that penalized likelihood using relative difference penalty, block sequential regularized expectation maximization (BSREM), achieves more accurate lesion quantitation than OSEM, importantly, maintains acceptable visual quality in...
High resolution PET imaging is being driven by recent advances in both scanner hardware and image reconstruction. As a result, images are reconstructed with smaller pixel sizes little or no post-reconstruction filtering. In certain circumstances, this combination can lead to distracting artifacts due high frequencies introduced the projectors. We present method that utilizes point spread function (PSF) processing mitigate these while preserving resolution. Our approach includes "hybrid-space...
Purpose: One major challenge facing simultaneous positron emission tomography (PET)/ magnetic resonance imaging (MRI) is PET attenuation correction (AC) measurement and evaluation of its accuracy. There a crucial need for the current emergent AC methodologies in terms absolute quantitative accuracy reconstructed images. Approach: To address this need, we developed evaluated lesion insertion tool PET/MRI that will facilitate process. This was Biograph mMR using phantom patient data. Contrast...
Model observers that replicate human are useful tools for assessing image quality based on detection tasks. Linear model including nonprewhitening matched filters (NPWMFs) and channelized Hotelling (CHOs) have been widely studied applied successfully to evaluate optimize performance. However, there is still room improvement in predicting observer responses In this study, we used a convolutional neural network predict two-alternative forced choice (2AFC) task PET imaging. Lesion-absent...
Positron emission tomography and magnetic resonance imaging (PET/MRI) scanners cannot be qualified in the manner adopted for hybrid PET computed (CT) devices. The main hurdle with qualification PET/MRI is that attenuation correction (AC) adequately measured conventional phantoms due to difficulty converting MRI images of physical structures (e.g., plastic) into electron density maps. Over last decade, a plethora novel MR-based algorithms have been developed more accurately derive properties...
An image reconstruction algorithm for rectangular PET scanner was developed using distance-driven projections. The geometrical sensitivity of the tubes response in geometry derived under assumption that detector area is very small relative to distances between detectors. Linograms were used histograms coincidences parallel panels, and “orthograms” proposed orthogonal panels. Open Gate simulate a box-shaped projections hot rod flood phantoms. reconstructed images from simulated demonstrated...
A major remaining challenge for magnetic resonance-based attenuation correction methods (MRAC) is their susceptibility to sources of MRI artifacts (e.g. implants, motion) and uncertainties due the limitations contrast accurate bone delineation density, separation air/bone). We propose using a Bayesian deep convolutional neural network that, in addition generating an initial pseudo-CT from MR data, also produces uncertainty estimates quantify data. These outputs are combined with MLAA...
A challenge in multicenter trials that use quantitative positron emission tomography (PET) imaging is the often unknown variability PET image values, typically measured as standardized uptake introduced by intersite differences global and resolution-dependent biases. We present a method for simultaneous monitoring of scanner calibration reconstructed resolution on per-scan basis using PET/computed (CT) "pocket" phantom. simulation phantom studies to optimize design construction PET/CT pocket...
Software for Tomographic Image Reconstruction (STIR: http://stir.sf.net) is an open source C++ library available reconstruction of emission tomography data. This work aims at the incorporation GE SIGNA PET/MR scanner in STIR and enables PET image with data corrections. The extracted from after acquisition includes a list raw files (emission, normalisation, geometric well counter calibration (wcc) factors), magnetic resonance attenuation correction (MRAC) images scanner-based reconstructions....