Search

PURPOSE Routine quarterly quality assurance (QA) assessment of single photon emission computed tomography (SPECT) systems includes analysis of multipurpose phantoms containing spheres and rods of various sizes. When evaluated by accreditation agencies, criteria applied to assess image quality are largely subjective. Determining a quantified image characteristic metric that emulates human reader impressions of image quality could be quite useful. Our investigation was conducted to ascertain whether image texture analysis metrics, such as those applied to PET scans to detect neoplasms, could prove helpful in linking qualitative statements of phantom sphere and rod visibility to quantified parameters. Because it is not obvious whether it is preferable to submit reconstructions to accrediting agencies performed using typical clinical (CLIN) protocol processing parameters or to follow agencies' filtered backprojection (FBP) suggestions, we applied texture analysis metrics to determine the degree to which these choices affect equipment capability assessment. METHODS AND MATERIALS Data were processed retrospectively for 125 different Tc-99 m SPECT scans of standardized phantoms for 14 rotating Anger detector systems as part of routine quarterly QA. Algorithms were written to compute several classes of image metrics: quantile curve metrics, image texture analysis gray-level co-occurrence matrix (GLCM) metrics, contrast metrics, and count histogram metrics. For qualitative image scores, two experienced physicists independently graded sphere and rod visibility on a 5-level scale and assigned dichotomous visibility scores, without knowledge of quantified texture analysis metrics or each other's readings. The same phantom was used to collect 15 additional data sets with two dual-detector SPECT/CT systems, reconstructed both by FBP parameters that have been suggested by accrediting agencies and by manufacturers' default settings for CLIN SPECT/CT bone imaging protocols by ordered subsets expectation maximization (OSEM), incorporating attenuation correction using the CT scan. Image characteristics metrics were compared for FBP and CLIN reconstructions. RESULTS For spheres, the metric with the strongest rank correlation with 5-level scale readings was the quantile curve slope (ρ = 0.83, P

Purpose: To investigate the 3D geometric distortion of four potential MR sequences for radiotheraptic applications, and its dependency on sequence-type, acquisition-orientation and receiver-bandwidth from a dedicated 1.5T 700mm-wide bore MR-simulator (Magnetom-Aera, Sienmens Healthcare, Erlangen, Germany), using a large customized geometric accuracy phantom. Methods: This work studied 3D gradient-echo (VIBE) and spin-echo (SPACE) sequences for anatomical imaging; a specific ultra-short-TE sequence (PETRA) potentially for bone imaging and MR-based dosimetry; and a motion-insensitive sequence (BLADE) for dynamic applications like 4D-MRI. Integrated geometric-correction was employed, three orthogonal acquisition-orientations and up to three receiver-bandwidths were used, yielding 27 acquisitions for testing (Table 1a).A customized geometric accuracy phantom (polyurethane, MR/CT invisible, W×L×H:55×55×32.5cm3) was constructed and filled with 3892 spherical markers (6mm diameter, MR/CT visible) arranged on a 25mm-interval 3D isotropic-grid (Fig.1). The marker positions in MR images were quantitatively calculated and compared against those in the CT-reference using customized MatLab scripts. Results: The average distortion within various diameter-of-spherical-volumes (DSVs) and the usable DSVs under various distortion limits were measured (Tables 1b-c). It was observed that distortions fluctuated when sequence-type, acquisition-orientation or receiver-bandwidth changed (e.g. within 300mm-DSV, the lowest/highest average distortions of VIBE were 0.40mm/0.59mm, a 47.5% difference). According to AAPM-TG66 (

Purpose: Clinical use of 18F-Sodium Fluoride (NaF) PET/CT in metastatic settings often lacks technology to quantitatively measure full disease dynamics due to high tumor burden. This study assesses radiomics-based extraction of NaF PET/CT measures, including global metrics of overall burden and local metrics of disease heterogeneity, in metastatic prostate cancer for correlation to clinical outcomes. Methods: Fifty-six metastatic Castrate-Resistant Prostate Cancer (mCRPC) patients had NaF PET/CT scans performed at baseline and three cycles into chemotherapy (N=16) or androgen-receptor (AR) inhibitors (N=39). A novel technology, Quantitative Total Bone Imaging (QTBI), was used for analysis. Employing hybrid PET/CT segmentation and articulated skeletal-registration, QTBI allows for response assessment of individual lesions. Various SUV metrics were extracted from each lesion (iSUV). Global metrics were extracted from composite lesion-level statistics for each patient (pSUV). Proportion of detected lesions and those with significant response (%-increase or %-decrease) was calculated for each patient based on test-retest limits for iSUV metrics. Cox proportional hazard regression analyses were conducted between imaging metrics and progression-free survival (PFS). Results: Functional burden (pSUVtotal) assessed mid-treatment was the strongest univariate predictor of PFS (HR=2.03; p

Purpose: Although the anatomy of interest should be positioned as close as possible to the isocenter of CT scanners, off-centering may be inevitable during certain exams in clinical practice such as lumbar spine and elbow imaging. Off-centering degrades image sharpness, generates streak artifacts, and sometimes creates blooming artifacts due to truncation. The purpose of this work was to investigate whether the use of model-based image reconstruction (MBIR) can alleviate the negative impacts of off-centering to achieve high quality CT bone imaging. Methods: Both an anthropomorphic phantom and an ex vivo swine elbow sample were scanned at centered and off-centered positions using clinical CT bone scan protocols. The magnitude of off-centering was determined from localizer radiographs. Both FBP and MBIR reconstructions were performed. For FBP, both standard and Bone Plus kernels commonly used in bone imaging were used. Objective assessment of image sharpness, noise standard deviation, and noise nonuniformity were performed. Additionally, we retrospectively analyzed human subject data acquired under off-centered conditions as a validation study. Results: In FBP images of the phantom, off-centering by 10 cm led to a 14% increase in noise (p

Purpose: To demonstrate how both the positioning of an object in the scan field of view (SFOV) and the size of the reconstructed field of view (RFOV) can both significantly affect and degrade the image sharpness. Methods: Clinical CT scans of the shoulders and elbows were inspected for image sharpness. GE scanners were used for the imaging with a “Bone Plus” reconstruction algorithm. In all scans the anatomy of interest was significantly displaced from the center of the SFOV (away from the scanner iso-center). Radial and azimuthal image sharpness measurements using the Modulation Transfer Function (MTF) were made using the same RFOV at the center of the SFOV and at varying distances from the center. These MTF measurements were also made with varying values of the RFOV. Results: Clinical images with the bone anatomy of interest positioned farther from the center of the SFOV showed increasing image softness occasionally to the point of being non-diagnostic. The MTF measurements showed that the resolution degraded significantly, by more than a factor of 2, moving from the center toward the periphery of the SFOV: from 12 cycles/cm to less than 6 cycles/cm at distances of 15 cm or more from the center of the SFOV. Increasing the RFOV beyond 20 cm also degraded the resolution, though less significantly. Conclusion: When preservation of bone detail is important one must not only use the proper bone reconstruction algorithm, but also must keep the body part as close to the center of the SFOV as possible. Though this may be difficult with some imaging tasks, such as imaging of the shoulder and elbow, every effort should be made to this end. For the best resolution, the RFOV should also be kept small, ideally to 20 cm or less. Partial research funding from GE HealthCare.

Purpose: MRI only radiation therapy has been increasingly gaining research interests. Conventionally, MRIs are fused with CT to take advantage of its superior soft tissue contrast. Such an approach has drawbacks such as additional uncertainty from fusion and extra cost. To achieve MR‐based simulation for radiation therapy, bone imaging is an important challenge that needs to be overcome. The ultra‐short T2 and T1 of cortical bone Result in low signal intensity, which presents difficulty for both dose calculation and patient setup in terms of digitally reconstructed radiograph (DRR) generation. Current solutions will either require manual bone contouring or multiple MR scans. We present a technique to generate brain DRR using bone MRI with a single Ultra Short Echo Time (UTE) sequence. Methods: Healthy volunteers were scanned at 1.5 Tesla using a radial UTE sequence. The sequence acquires two images at two different echo times (0.07ms and 3.6ms). The two images were processed using in‐house software, which performs pixel‐by‐pixel magnitude subtraction of the two images. The difference image was subsequently divided pixel‐by‐pixel to the long TE (3.6ms) image. As the cortical bone has larger percentile difference between the UTE and long TE image compared with other tissues, the bone is bright in the processed image. The resultant bone images were subsequently loaded into commercial treatment planning system to generate DRRs. Results: The majority of cranial, facial and vertebral bones are well visualized in both volunteers. The fused image of the long TE image with the processed bone image demonstrates the accuracy of automatic bone identification using our technique. The generated DRR is of acceptable quality. Conclusion: This study shows the potential of DRR generation with single MR sequence. Further work will be needed on MR sequence development and post‐processing procedure to achieve MR bone image with comparable quality to CT.

Purpose: Prostate cancer is the most prevalent cancer among males in the US, and over 90% of patients develop bone metastases. Current treatment response evaluation metrics, such as RECIST, are not suited for assessing bone lesions. The aim of this work is to present an automated segmentation method that identifies and characterizes bone lesions and determine its sensitivity to thresholding. Methods: Patients with metastatic prostate cancer were treated with molecular target therapy and received whole body [F‐18]NaF (bone imaging) PET/CT scans during the course of treatment. Scans were evaluated using the Quantitative Total Bone Imaging (QTBI) methodology, which assesses total bone and individual bone lesions in patients with bone metastases. Individual lesions were segmented using thresholds ranging from 20% to 80% of the maximum standardized uptake value (SUV) in bone. The mean SUV, total SUV, and volume of the total bone and individual lesions were determined as well as the number of lesions. Treatment response was assessed as the percent change in the imaging metrics relative to pre‐treatment. Results: The QTBI methodology was able to identify and characterize bone lesions. Of the treatment response metrics, change in number of lesions showed the greatest dependence on the threshold. The standard deviation of patient responses ranged from 13.5 to 34.5. Change in mean SUV had the least sensitivity to threshold (SD=0.5–2.0). Change in total SUV and volume both showed intermediate variation (SD=7.5–11.5 and 11.0–12.5, respectively). Conclusions: The QTBI methodology is able to identify and characterize bone lesions. However, because most imaging metric responses vary with different thresholds, careful selection of SUV threshold for segmentation is necessary to accurately identify and characterize bone lesions and assess treatment response. With further refinement, this methodology can be a clinically useful tool to improve patient care and test new therapies for use in metastatic bone disease.

Objective 18F‐Fluorodeoxyglucose (FDG) is a metabolic activity marker with limited use for skeletal lesions. Sodium18F‐flouride (NaF) is a bone imaging agent that complements the FDG limitations for skeletal lesions. The kinetic analysis of combined NaF/FDG PET scan promises concurrent determination of both tumor metabolism and vasculature parameters as well as successful separation of each tracer uptake. The aim of this work is to optimize the combined NaF/FDG PET scanning protocol in order to maximize accuracy in kinetic model estimation of the tumor's parameters. Materials and Methods Simulations assessed the feasibility and accuracy of kinetic analysis of combined NaF/FDG PETimages. Tissue time activity curves (tTAC) were generated with randomized parameters and PET voxel‐wise noise level. Various time delays between NaF and FDG injection times were simulated while keeping the total scan duration at 60 minutes. Parameters were estimated with two simplified versions of the model used for tTAC's generation. Estimated parameters were compared against the simulated parameters. The optimized sequence was tested on canine NaF/FDG PET data.Results Parameter estimation accuracy was maximized if the NaF was injected first followed by the FDG injection of at least 40 minutes later. The estimation accuracy for k1 parameters ranged from 15% to 20% for FDG and 25% to 30% for NaF. Ki estimation accuracy ranged from 40% to 55% for FDG and 10% to 25% for NaF. Vasculature fraction estimation accuracy ranged from 20% to 40%. Results of clinical data kinetic analysis shows that both k1 parameters are highly correlated which implies that it might be common for both radiopharmaceuticals in our model._Conclusions This study has shown that kinetic analysis of combined NaF/FDG PETimages is feasible using an optimized injection and scanning protocol. For the given cases we found the optimal scanning protocol and quantified parameter estimation accuracy.