Deep learning for improving PET/CT attenuation correction by elastic registration of anatomical data.

BACKGROUND: For PET/CT, the CT transmission data are used to correct the PET emission data for attenuation. However, subject motion between the consecutive scans can cause problems for the PET reconstruction. A method to match the CT to the PET would reduce resulting artifacts in the reconstructed images. PURPOSE: This work presents a deep learning technique for inter-modality, elastic registration of PET/CT images for improving PET attenuation correction (AC). The feasibility of the technique is demonstrated for two applications: general whole-body (WB) imaging and cardiac myocardial perfusion imaging (MPI), with a specific focus on respiratory and gross voluntary motion. MATERIALS AND METHODS: A convolutional neural network (CNN) was developed and trained for the registration task, comprising two distinct modules: a feature extractor and a displacement vector field (DVF) regressor. It took as input a non-attenuation-corrected PET/CT image pair and returned the relative DVF between them-it was trained in a supervised fashion using simulated inter-image motion. The 3D motion fields produced by the network were used to resample the CT image volumes, elastically warping them to spatially match the corresponding PET distributions. Performance of the algorithm was evaluated in different independent sets of WB clinical subject data: for recovering deliberate misregistrations imposed in motion-free PET/CT pairs and for improving reconstruction artifacts in cases with actual subject motion. The efficacy of this technique is also demonstrated for improving PET AC in cardiac MPI applications. RESULTS: A single registration network was found to be capable of handling a variety of PET tracers. It demonstrated state-of-the-art performance in the PET/CT registration task and was able to significantly reduce the effects of simulated motion imposed in motion-free, clinical data. Registering the CT to the PET distribution was also found to reduce various types of AC artifacts in the reconstructed PET images of subjects with actual motion. In particular, liver uniformity was improved in the subjects with significant observable respiratory motion. For MPI, the proposed approach yielded advantages for correcting artifacts in myocardial activity quantification and potentially for reducing the rate of the associated diagnostic errors. CONCLUSION: This study demonstrated the feasibility of using deep learning for registering the anatomical image to improve AC in clinical PET/CT reconstruction. Most notably, this improved common respiratory artifacts occurring near the lung/liver border, misalignment artifacts due to gross voluntary motion, and quantification errors in cardiac PET imaging.

Change in positron emission tomography perfusion imaging quality with a data-driven motion correction algorithm.

INTRODUCTION: Cardiac motion frequently reduces the interpretability of PET images. This study utilized a prototype data-driven motion correction (DDMC) algorithm to generate corrected images and compare DDMC images with non-corrected images (NMC) to evaluate image quality and change of perfusion defect size and severity. METHODS: Rest and stress images with NMC and DDMC from 40 consecutive patients with motion were rated by 2 blinded investigators on a 4-point visual ordinal scale (0: minimal motion; 1: mild motion; 2: moderate motion; 3: severe motion/uninterpretable). Motion was also quantified using Dwell Fraction, which is the fraction of time the motion vector shows the heart to be within 6 mm of the corrected position and was derived from listmode data of NMC images. RESULTS: Minimal motion was seen in 15% of patients, while 40%, 30%, and 15% of patients had mild moderate and severe motion, respectively. All corrected images showed an improvement in quality and were interpretable after processing. This was confirmed by a significant correlation (Spearman's correlation coefficient 0.626, P < .001) between machine measurement of motion quantification and physician interpretation. CONCLUSION: The novel DDMC algorithm improved quality of cardiac PET images with motion. Correlation between machine measurement of motion quantification and physician interpretation was significant.

A preliminary evaluation of a high temporal resolution data-driven motion correction algorithm for rubidium-82 on a SiPM PET-CT system.

BACKGROUND: In myocardial perfusion PET, images are acquired during vasodilator stress, increasing the likelihood of intra-frame motion blurring of the heart in reconstructed static images to assess relative perfusion. This work evaluated a prototype data-driven motion correction (DDMC) algorithm designed specifically for cardiac PET. METHODS: A cardiac torso phantom, with a solid defect, was scanned stationary and being manually pulled to-and-fro in the axial direction with a random motion. Non-motion-corrected (NMC) and DDMC images were reconstructed. Total perfusion deficit was measured in the defect and profiles through the cardiac insert were defined. In addition, 46 static perfusion images from 36 rubidium-82 MPI patients were selected based upon a perception of motion blurring in the images. NMC and DDMC images were reconstructed, blinded, and scored on image quality and perceived motion. RESULTS: Phantom data demonstrated near-perfect recovery of myocardial wall visualization and defect quantification with DDMC compared with the stationary phantom. Quality of clinical images was NMC: 10 non-diagnostic, 31 adequate, and 5 good; DDMC images: 0 non-diagnostic, 6 adequate, and 40 good. CONCLUSION: The DDMC algorithm shows great promise in rubidium MPI PET with substantial improvements in image quality and the potential to salvage images considered non-diagnostic due to significant motion blurring.

The prevalence of image degradation due to motion in rest-stress rubidium-82 imaging on a SiPM PET-CT system.

BACKGROUND: Motion of the heart is known to affect image quality in cardiac PET. The prevalence of motion blurring in routine cardiac PET is not fully appreciated due to challenges identifying subtle motion artefacts. This study utilizes a recent prototype Data-Driven Motion Correction (DDMC) algorithm to generate corrected images that are compared with non-corrected images to identify visual differences in relative rubidium-82 perfusion images due to motion. METHODS: 300 stress and 300 rest static images were reconstructed with DDMC and without correction (NMC). The 600 DDMC/NMC image pairs were assigned Visual Difference Score (VDS). The number of non-diagnostic images were noted. A "Dwell Fraction" (DF) was derived from the data to quantify motion and predict image degradation. RESULTS: Motion degradation (VDS = 1 or 2) was evident in 58% of stress images and 33% of rest images. Seven NMC images were non-diagnostic-these originated from six studies giving a 2% rate of non-diagnostic studies due to motion. The DF metric was able to effectively predict image degradation. The DDMC heart identification and tracking was successful in all images. CONCLUSION: Motion degradation is present in almost half of all relative perfusion images. The DDMC algorithm is a robust tool for predicting, assessing and correcting image degradation.

Assessment of motion correction in dynamic rubidium-82 cardiac PET with and without frame-by-frame adjustment of attenuation maps for calculation of myocardial blood flow.

BACKGROUND: Patient motion during pharmacological stressing can have substantial impact on myocardial blood flow (MBF) estimated from dynamic PET. This work evaluated a motion correction algorithm with and without adjustment of the PET attenuation map. METHODS: Frame-by-frame motion correction was performed by three users on 30 rubidium-82 studies. Data were divided equally into three groups of motion severity [mild (M1), moderate (M2) and severe (M3)]. MBF data were compared for non-motion corrected (NC), motion-corrected-only (MC) and with adjustment of the attenuation map (MCAC). Percentage differences of MBF were calculated in the coronary territories and 17-segment polar plots. Polar plots of spill-over were also generated from the data. RESULTS: Median differences of 23% were seen in the RCA and 18% for the LAD in the M3 category for MC vs NC images. Differences for MCAC vs MC images were considerably smaller and typically < 10%. Spill-over plots for MC and MCAC were notably more uniform compared with NC images. CONCLUSION: Motion correction for dynamic rubidium data is desirable for future MBF software updates. Adjustment of the PET attenuation map results in only marginal differences and therefore is unlikely to be an essential requirement. Assessing the uniformity of spill-over plots is a useful visual aid for verifying motion correction techniques.

Assessing Reliability of Myocardial Blood Flow After Motion Correction With Dynamic PET Using a Bayesian Framework.

The estimation of myocardial blood flow (MBF) in dynamic PET can be biased by many different processes. A major source of error, particularly in clinical applications, is patient motion. Patient motion, or gross motion, creates displacements between different PET frames as well as between the PET frames and the CT-derived attenuation map, leading to errors in MBF calculation from voxel time series. Motion correction techniques are challenging to evaluate quantitatively and the impact on MBF reliability is not fully understood. Most metrics, such as signal-to-noise ratio (SNR), are characteristic of static images, and are not specific to motion correction in dynamic data. This study presents a new approach of estimating motion correction quality in dynamic cardiac PET imaging. It relies on calculating a MBF surrogate, K1 , along with the uncertainty on the parameter. This technique exploits a Bayesian framework, representing the kinetic parameters as a probability distribution, from which the uncertainty measures can be extracted. If the uncertainty extracted is high, the parameter studied is considered to have high variability - or low confidence - and vice versa. The robustness of the framework is evaluated on simulated time activity curves to ensure that the uncertainties are consistently estimated at the multiple levels of noise. Our framework is applied on 40 patient datasets, divided in 4 motion magnitude categories. Experienced observers manually realigned clinical datasets with 3D translations to correct for motion. K1 uncertainties were compared before and after correction. A reduction of uncertainty after motion correction of up to 60% demonstrates the benefit of motion correction in dynamic PET and as well as provides evidence of the usefulness of the new method presented.

Evaluation of MLACF based calculated attenuation brain PET imaging for FDG patient studies.

Calculating attenuation correction for brain PET imaging rather than using CT presents opportunities for low radiation dose applications such as pediatric imaging and serial scans to monitor disease progression. Our goal is to evaluate the iterative time-of-flight based maximum-likelihood activity and attenuation correction factors estimation (MLACF) method for clinical FDG brain PET imaging. FDG PET/CT brain studies were performed in 57 patients using the Biograph mCT (Siemens) four-ring scanner. The time-of-flight PET sinograms were acquired using the standard clinical protocol consisting of a CT scan followed by 10 min of single-bed PET acquisition. Images were reconstructed using CT-based attenuation correction (CTAC) and used as a gold standard for comparison. Two methods were compared with respect to CTAC: a calculated brain attenuation correction (CBAC) and MLACF based PET reconstruction. Plane-by-plane scaling was performed for MLACF images in order to fix the variable axial scaling observed. The noise structure of the MLACF images was different compared to those obtained using CTAC and the reconstruction required a higher number of iterations to obtain comparable image quality. To analyze the pooled data, each dataset was registered to a standard template and standard regions of interest were extracted. An SUVr analysis of the brain regions of interest showed that CBAC and MLACF were each well correlated with CTAC SUVrs. A plane-by-plane error analysis indicated that there were local differences for both CBAC and MLACF images with respect to CTAC. Mean relative error in the standard regions of interest was less than 5% for both methods and the mean absolute relative errors for both methods were similar (3.4% ± 3.1% for CBAC and 3.5% ± 3.1% for MLACF). However, the MLACF method recovered activity adjoining the frontal sinus regions more accurately than CBAC method. The use of plane-by-plane scaling of MLACF images was found to be a crucial step in order to obtain improved activity estimates. Presence of local errors in both MLACF and CBAC based reconstructions would require the use of a normal database for clinical assessment. However, further work is required in order to assess the clinical advantage of MLACF over CBAC based method.

BUY QUIET INITIATIVE IN THE USA.

Noise-induced hearing loss is still considered one of the most common work-related illnesses in the United States of America. The U.S. National Institute for Occupational Safety and Health launched a national Buy Quiet campaign to raise awareness of the importance of purchasing quieter equipment. Buy Quiet encourages companies to seek out and demand quieter equipment thus driving the market to design and create quieter products. In the long run, investment in noise controls should be more prevalent as the market demands quieter products. This paradigm occurs as the market for quieter products expands both from the supply side (manufacturers) and the demand side (tool and equipment purchasers). The key to experiencing the reduced costs and increased benefits of Buy Quiet will be to develop partnerships between manufacturers and consumers. To this end, the U.S. National Institute for Occupational Safety and Health continues to work with partners to educate stakeholders about the risks and true costs of noise-induced hearing loss, as well as the economic benefits of buying quieter equipment.

Development of an empirical model to aid in designing airborne infection isolation rooms.

Airborne infection isolation rooms (AIIRs) house patients with tuberculosis, severe acute respiratory syndrome (SARS), and many other airborne infectious diseases. Currently, facility engineers and designers of heating, ventilation, and air-conditioning (HVAC) systems have few analytical tools to estimate a room's leakage area and establish an appropriate flow differential (DeltaQ) in hospitals, shelters, and other facilities where communicable diseases are present. An accurate estimate of leakage area and selection of DeltaQ is essential for ensuring that there is negative pressure (i.e., pressure differential [DeltaP]) between an AIIR and adjoining areas. National Institute for Occupational Safety and Health (NIOSH) researchers evaluated the relationship between DeltaQ and DeltaP in 67 AIIRs across the United States and in simulated AIIR. Data gathered in the simulated AIIR was used to develop an empirical model describing the relationship between DeltaQ, DeltaP, and leakage area. Data collected in health care facilities showed that the model accurately predicted the leakage area 44 of 48 times. Statistical analysis of the model and experimental validation showed that the model effectively estimated the actual leakage area from -39% to +22% with 90% confidence. The NIOSH model is an effective, cost-cutting tool that can be used by HVAC engineers and designers to estimate leakage area and select an appropriate DeltaQ in AIIRs to reduce the airborne transmission of disease.

Noise exposure assessment and abatement strategies at an indoor firing range.

Exposure to hazardous impulse noise is common during the firing of weapons at indoor firing ranges. The aims of this study were to characterize the impulse noise environment at a law enforcement firing range; document the insufficiencies found at the range from a health and safety standpoint; and provide noise abatement recommendations to reduce the overall health hazard to the auditory system. Ten shooters conducted a typical live-fire exercise using three different weapons--the Beretta.40 caliber pistol, the Remington.308 caliber shotgun, and the M4.223 caliber assault rifle. Measurements were obtained at 12 different positions throughout the firing range and adjacent areas using dosimeters and sound level meters. Personal and area measurements were recorded to a digital audio tape (DAT) recorder for further spectral analysis. Peak pressure levels inside the firing range reached 163 decibels (dB) in peak pressure. Equivalent sound levels (Leq) ranged from 78 decibels, A-weighted (dBA), in office area adjacent to the range to 122 dBA inside the range. Noise reductions from wall structures ranged from 29-44 dB. Noise abatement strategies ranged from simple noise control measures (such as sealing construction joints and leaks) to elaborate design modifications to eliminate structural-borne sounds using acoustical treatments. Further studies are needed to better characterize the effects of firing weapons in enclosed spaces on hearing and health in general.