Clinical implications of different image reconstruction parameters for interpretation of whole-body PET studies in cancer patients.

The standardized uptake value (SUV) is the most commonly used parameter to quantify the intensity of radiotracer uptake in tumors. Previous studies suggested that measurements of (18)F-FDG accumulation in tissue might be affected by the image reconstruction method, but the clinical relevance of these findings has not been assessed. METHODS: Phantom studies were performed and clinical whole-body (18)F-FDG PET images of 85 cancer patients were analyzed. All images were reconstructed using either filtered backprojection (FBP) with measured attenuation correction (MAC) or iterative reconstruction (IR) with segmented attenuation correction (SAC). In a subset of 15 patients, images were reconstructed using all 4 combinations of IR+SAC, IR+MAC, FBP+SAC, and FBP+MAC. For phantom studies, a sphere containing (18)F-FDG was placed in a water-filled cylinder and the activity concentration of that sphere was measured in FBP and IR reconstructed images using all 4 combinations. Clinical studies were displayed simultaneously and identical regions of interest (ROIs, 50 pixels) were placed in liver, urinary bladder, and tumor tissue in both image sets. SUV max (maximal counts per pixel in ROI) and SUV avg (average counts per pixel) were measured. RESULTS: In phantom studies, measurements from FBP images underestimated the true activity concentration to a greater degree than those from IR images (20% vs. 5% underestimation). In patient studies, SUV derived from FBP images were consistently lower than those from IR images in both normal and tumor tissue: Tumor SUV max with IR+SAC was 9.6 +/- 4.5, with IR+MAC it was 7.7 +/- 3.5, with FBP+MAC it was 6.9 +/- 3.0, and with FBP+SAC it was 8.6 +/- 4.1 (all P < 0.01 vs. IR+SAC). Compared with IR+SAC, SUV from FBP+MAC images were 25%-30% lower. Similar discrepancies were noted for liver and bladder. Discrepancies between measurements became more apparent with increasing (18)F-FDG concentration in tissue. CONCLUSION: SUV measurements in whole-body PET studies are affected by the applied methods for both image reconstruction and attenuation correction. This should be considered when serial PET studies are done in cancer patients. Moreover, if SUV is used for tissue characterization, different cutoff values should be applied, depending on the chosen method for image reconstruction and attenuation correction.     

Evaluation of iterative reconstruction (OSEM) versus filtered back-projection for the assessment of myocardial glucose uptake and myocardial perfusion using dynamic PET

Purpose Iterative reconstruction methods based on ordered-subset expectation maximisation (OSEM) has replaced filtered backprojection (FBP) in many clinical settings owing to the superior image quality. Whether OSEM is as accurate as FBP in quantitative positron emission tomography (PET) is uncertain. We compared the accuracy of OSEM and FBP for regional myocardial ¹⁸F-FDG uptake and ¹³NH₃ perfusion measurements in cardiac PET. Methods Ten healthy volunteers were studied. Five underwent dynamic ¹⁸F-FDG PET during hyperinsulinaemic–euglycaemic clamp, and five underwent ¹³NH₃ perfusion measurement during rest and adenosine-induced hyperaemia. Images were reconstructed using FBP and OSEM ± an 8-mm Gaussian post-reconstruction filter. Results Filtered and unfiltered images showed agreement between the reconstruction methods within ±2SD in Bland-Altman plots of K _i values. The use of a Gaussian filter resulted in a systematic underestimation of K _i in the filtered images of 11%. The mean deviation between the reconstruction methods for both unfiltered and filtered images was 1.3%. Agreement within ±2SD between the methods was demonstrated for perfusion rate constants up to 2.5 min⁻¹, corresponding to a perfusion of 3.4 ml g⁻¹ min⁻¹. The mean deviation between the two methods for unfiltered data was 2.7%, and for filtered data, 5.3%. Conclusion The ¹⁸F-FDG uptake rate constants showed excellent agreement between the two reconstruction methods. In the perfusion range up to 3.4 ml g⁻¹ min⁻¹, agreement between ¹³NH₃ perfusion obtained with OSEM and FBP was acceptable. The use of OSEM for measurement of perfusion values higher than 3.4 ml g⁻¹ min⁻¹ requires further evaluation.     

Quantitative comparison of analytic and iterative reconstruction methods in 2- and 3-dimensional dynamic cardiac 18F-FDG PET.

The aim of this work was to compare the quantitative accuracy of iteratively reconstructed cardiac (18)F-FDG PET with that of filtered backprojection for both 2-dimensional (2D) and 3-dimensional (3D) acquisitions and to establish an optimal procedure for imaging myocardial viability with (18)F-FDG PET. METHODS: Eight patients underwent dynamic cardiac (18)F-FDG PET using an interleaved 2D/3D scan protocol, enabling comparison of 2D and 3D acquisitions within the same patient and study. A 10-min transmission scan was followed by a 10-min, 25-frame dynamic 3D scan and then by a series of 10 alternating 5-min 3D and 2D scans. Images were reconstructed with filtered backprojection (FBP) or attenuation-weighted ordered-subsets expectation maximization (OSEM), combined with Fourier rebinning (FORE) for 3D acquisitions, applying all usual corrections. Regions of interest (ROIs) were drawn in the myocardium, left ventricle, and ascending aorta, with the last 2 being used to define image-derived input functions (IDIFs). Patlak graphical analysis was used to compare net (18)F-FDG uptake in the myocardium, calculated from either 2D or 3D data, after reconstruction with FBP or OSEM. Either IDIFs or arterial sampling was used as the input function. The same analysis was performed on parametric images. RESULTS: A good correlation (r(2) > 0.99) was found between net (18)F-FDG uptake values for a myocardium ROI determined using each acquisition and reconstruction method and blood-sampling input functions. A similar result was found for parametric images. The ascending aorta was the best choice for IDIF definition. CONCLUSION: Good correlation and no bias of net (18)F-FDG uptake in relation to that based on FBP images, combined with less image noise, make 3D acquisition with FORE plus attenuation-weighted OSEM reconstruction the preferred choice for cardiac (18)F-FDG PET studies.     

Image reconstruction using filtered backprojection and iterative method: effect on motion artifacts in myocardial perfusion SPECT

Patient motion during myocardial perfusion SPECT is a common source of errors. The extent and severity of motion artifacts have been described for filtered backprojection (FBP) reconstruction. In recent years, iterative reconstruction has been used increasingly in reconstruction of myocardial perfusion SPECT images and has been shown to be more accurate than FBP even in cases of incomplete datasets. This study evaluated the effect of iterative reconstruction on the extent and severity of motion artifacts. METHODS: Six normal, motion-free, and nongated (99m)Tc myocardial perfusion SPECT scans were selected, and simulated motion of 3 pixels was applied to the early, middle, and late phases of acquisition in 2 types of movement, returning and nonreturning. The images were acquired by a single-head gamma-camera in 32 steps at 30 s per step and in a 180 degrees arc from right anterior oblique to left posterior oblique. All original and shifted images were reconstructed using FBP and ordered-subset expectation maximization (OSEM) techniques and interpreted by 2 nuclear medicine specialists qualitatively and semiquantitatively (using 17 segments and a 5-point scoring system). RESULTS: Overall, 68.1% and 70.8% of shifted images were categorized as definitely abnormal in the FBP and OSEM reconstructions, respectively (P > 0.5). The mean summed score was 11.9 (+/-5.7) and 11.3 (+/-5.2) for nonreturning shifted images (P = 0.13) and 5.2 (+/-2.4) and 3.9 (+/-2.0) for returning shifted images (P < 0.001) in the OSEM and FBP reconstructions, respectively. The incidence of defects in different myocardial segments was similar with the 2 reconstruction methods. The summed score was higher with shifting in the middle phase of acquisition than in the late or early phase. CONCLUSION: Our study showed that the incidence of abnormal findings and the location of defects were not different between the 2 reconstruction types; however, with semiquantitative assessment, the severity of defects increased with OSEM reconstruction. Although OSEM reconstruction has been reported to be more tolerant to missing data than is FBP reconstruction, our study showed that OSEM reconstruction may be less tolerant to motion artifacts than is FBP reconstruction.     

Noise reduction in oncology FDG PET images by iterative reconstruction: a quantitative assessment.

Tumor detection depends on the contrast between tumor activity and background activity and on the image noise in these 2 regions. The lower the image noise, the easier the tumor detection. Tumor activity contrast is determined by physiology. Noise, however, is affected by many factors, including the choice of reconstruction algorithm. Previous simulation and phantom measurements indicated that the ordered-subset expectation maximization (OSEM) algorithm may produce less noisy images than does the usual filtered backprojection (FBP) method, at equivalent resolution. To see if this prediction would hold in actual clinical situations, we quantified noise in clinical images reconstructed with both OSEM and FBP. METHODS: Three patients (2 with colon cancer, 1 with breast cancer) were imaged with FDG PET using a "gated replicate" technique that permitted accurate measurement of noise at each pixel. Each static image was acquired as a gated image sequence, using a pulse generator with a 1-s period, yielding 40 replicate images over the 10- to 15-min imaging time. The images were or were not precorrected for attenuation and were reconstructed with both FBP and OSEM at comparable resolution. From these data, images of pixel mean, SD, and signal-to-noise ratio (S/N) could be produced, reflecting only noise caused by the statistical fluctuations in the emission process. RESULTS: Noise did not vary greatly over each FBP image, even when image intensity varied greatly from one region to the next, causing S/N to be worse in low-activity regions than in high-activity regions. In contrast, OSEM had high noise in hot regions and low noise in cold regions. OSEM had a much better S/N than did FBP in cold regions of the image, such as the lungs (in the attenuation-corrected images), where improvements in S/N averaged 160%. Improvements with OSEM were less dramatic in hotter areas such as the liver (averaging 25% improvement in the attenuation-corrected images). In very hot tumors, FBP actually produced higher S/Ns than did OSEM. CONCLUSION: We conclude that OSEM reconstruction can significantly reduce image noise, especially in relatively low-count regions. OSEM reconstruction failed to improve S/N in very hot tumors, in which S/N may already be adequate for tumor detection.     

SPECT图像重建方法对Hoffmann模型图像质量的影响

目的探讨SPECT显像不同断层重建方法对Hoffmann模型图像质量的影响。方法采用放射性线源及Hoffmann模型,进行SPECT配平行孔低能高分辨准直器断层采集。对线源图像和Hoffmann模型图像均用滤波反投影（FBP）法和有序子集最大期望值迭代（OSEM）法进行断层重建。对线源重建图像计算2种重建算法的半高宽（FWHM）值,视觉评价2种方法重建的Hoffmann模型图像,并比较2种重建方法的重建时间和模型特定感兴趣区（ROI）。采用SPSS15．0软件,2种重建方法的脑叶及基底节ROI与小脑ROI计数比值行两独立样本t检验。结果平行孔低能高分辨准直器采集的线源断层图像经FBP法和OSEM法重建,FWHM值分别为18．77mm,12．62mm。OSEM法重建所得Hoffmann模型总的图像质量、基底节区的显示以及核团分辨程度均优于FBP法。FBP法和OSEM法重建时间分别为80s和100s,均在临床允许范围之内。OSEM迭代与FBP法脑叶及基底节ROI与小脑ROI比值差异无统计学意义（t=-0．332,P=0．750）。结论在现有软硬件技术条件下,平行孔低能高分辨准直器配合OSEM法优于FBP法,可获得空间分辨率和图像质量较好的Hoffmann模型重建图像。     

Attenuation correction in whole-body FDG oncological studies: the role of statistical reconstruction.

Whole-body fluorine-18 fluoro-2-d-deoxyglucose positron emission tomography (FDG-PET) is widely used in clinical centres for diagnosis, staging and therapy monitoring in oncology. Images are usually not corrected for attenuation since filtered backprojection (FBP) reconstruction methods require a 10 to 15-min transmission scan per bed position on most current PET devices equipped with germanium-68 rod transmission sources. Such an acquisition protocol would increase the total scanning time beyond acceptable limits. The aim of this work is to validate the use of iterative reconstruction methods, on both transmission and emission scans, in order to obtain a fully corrected whole-body study within a reasonable scanning time of 60 min. Five minute emission and 3-min transmission scans are acquired at each of the seven bed positions. The transmission data are reconstructed with OSEM (ordered subsets expectation maximization) and the last iteration is reprojected to obtain consistent attenuation correction factors (ACFs). The emission image is then also reconstructed with OSEM, using the emission scan corrected for normalization, scatter and decay together with the set of consistent ACFs as inputs. The total processing time is about 35 min, which is acceptable in a clinical environment. The image quality, readability and accuracy of uptake quantification were assessed in 38 patients scanned for various malignancies. The sensitivity for tumour detection was the same for the non-attenuation-corrected (NAC-FBP) and the attenuation-corrected (AC-OSEM) images. The AC-OSEM images were less noisy and easier to interpret. The interobserver reproducibility was significantly increased when compared with non-corrected images (96.1% vs 81.1%, P<0.01). Standardized uptake values (SUVs) measured on images reconstructed with OSEM (AC-OSEM) and filtered backprojection (AC-FBP) were similar in all body regions except in the pelvic area, where SUVs were higher on AC-FBP images (mean increase 7.74%, P<0. 01). Our results show that, when statistical reconstruction is applied to both transmission and emission data, high quality quantitative whole-body images are obtained within a reasonable scanning (60 min) and processing time, making it applicable in clinical practice.     

不同处理方式对FDG PET图像质量的影响

目的：研究不同重建方式对全身18F－脱氧葡萄糖（FDG）PET图像质量的影响。方法：5例正常对照者和5例恶性肿瘤患者全身18F－FDG PET显像（皆同时进行发射和穿透扫描），所有受检者的原始资料分别采用多种重建方式（无衰减滤波法重建，衰减滤波法重建，衰减迭代法重建），同时分析不同衰减平滑方法（Nonquantitative 和Gaussian)对图像质量的影响。结果：无衰减校正影像存在身体中间组织衰减明显，组织结构显示不清，病变容易变形，伪影多，图像质量差，病灶定位较困难，不能定量计算标准取值（SUV）等缺点，衰减校正滤波法重建图像在显示组织结构，定位病变及定量计算方面较非常衰减法好，但图像质量仍较差，衰减校正迭代法重建图像质量好，组织结构显示清楚，伪影少，病灶无变形，定位及定量都较准确，明显优于衰减和非衰减校正滤波法，衰减平滑方法对滤波法影响大，用Nonquantitative法明显优于Gaussian法，衰减平滑方法对迭代法影响较小，Gaussian法比Nonquantitative法略好。结论：不同重建方式对18F－FDG PET图像质量影响很大，以衰减校正迭代法重建（衰减平滑方法为Gaussian)图像质量最好。     

Comparing filtered backprojection and ordered-subsets expectation maximization for small-lesion detection and localization in 67Ga SPECT.

Iterative reconstruction of SPECT images has recently become clinically available as an alternative to filtered backprojection (FBP). However, there is conflicting evidence on whether iterative reconstruction, such as with the ordered-subsets expectation maximization (OSEM) algorithm, improves diagnostic performance over FBP. The study objective was to determine if the detection and localization of small lesions in simulated thoracic gallium SPECT images are better with OSEM reconstruction than with FBP, both with and without attenuation correction (AC). METHODS: Images were simulated using an analytic projector acting on the mathematic cardiac torso computer phantom. Perfect scatter rejection was assumed. Lesion detection accuracy was assessed using localization receiver operating characteristic methodology. The images were read by 5 nuclear medicine physicians. For each reconstruction strategy and for each observer, data were collected in 2 viewing sessions of 100 images. Two-way ANOVA and, when indicated, the Scheffé multiple comparisons test were applied to check for significant differences. RESULTS: Little difference in the accuracy of detection or localization was seen between FBP with and without AC. OSEM with AC extended the contrast range for accurate lesion detection and localization over that of the other methods investigated. Without AC, no significant difference between OSEM and FBP reconstruction was detected. CONCLUSION: OSEM with AC may improve the detection and localization of thoracic gallium-labeled lesions over FBP reconstruction.     

Experimental and clinical evaluation of iterative reconstruction (OSEM) in dynamic PET: quantitative characteristics and effects on kinetic modeling.

The purpose of this study was to investigate the quantitative properties and effects of ordered-subset expectation maximization (OSEM) on kinetic modeling compared with filtered backprojection (FBP) in dynamic PET studies. Both phantom and patient studies were performed. METHODS: For phantom studies dynamic two-dimensional emission scans with 10-min frames and 20-min scan intervals were acquired over a 14-h period using an HR+ PET scanner. Various phantoms were scanned: 2-, 5-, 10-, and 20-cm-diameter phantoms filled with an 18F solution (300 kBq/mL) and a NEMA phantom filled with an 18F background (40 kBq/mL) and a cold or 11C insert (450 kBq/mL). Transmission (Tx) scans of 5-60 min were acquired. Data were reconstructed using FBP Hanning 0.5 and OSEM with 2-12 iterations and 12 or 24 subsets. Quantitative accuracy and noise characteristics were assessed. For patient studies, five cardiac, three oncologic, and three brain dynamic 18F-FDG scans were used. Five reconstructions were performed: FBP Hanning 0.5, and OSEM 2 x 12 and OSEM 4 x 16 with and without 5-mm full width at half maximum smoothing. Time-activity curves were calculated using volumes of interest. The input function was derived from arterial sampling. Metabolic rate of glucose (MRglu) was calculated with a standard two-tissue compartment model and Patlak analysis. RESULTS: Contribution of Tx noise to the reconstructed image was smaller for OSEM than for FBP. Differences in signal-to-noise ratio between FBP and OSEM depended on number of iterations and phantom size. Bias with OSEM was observed for regions enclosed within a 5- to 10-fold hotter background. For cardiac studies OSEM 2 x 12 and OSEM 4 x 16 resulted in 13% and 21% higher pixel values and 9% and 15% higher MRglu values compared with FBP. Smoothing decreased all these values to 2%. Similar results were found for most tumor studies. For brain studies MRglu of FBP and OSEM 4 x 16 agreed within 2%. Use of OSEM image-derived input functions for cardiac PET studies resulted in a decrease in calculated MRglu of about 15%. CONCLUSION: For most PET studies OSEM has equal quantitative accuracy as FBP. The higher pixel and MRglu values are explained by the better resolution of OSEM. However, OSEM does not provide accurate image-derived input functions for FDG cardiac PET studies because of bias in regions located within a hotter background.     

Attenuation correction in whole-body FDG oncological studies: the role of statistical reconstruction

Whole-body fluorine-18 fluoro-2-d-deoxyglucose positron emission tomography (FDG-PET) is widely used in clinical centres for diagnosis, staging and therapy monitoring in oncology. Images are usually not corrected for attenuation since filtered backprojection (FBP) reconstruction methods require a 10 to 15-min transmission scan per bed position on most current PET devices equipped with germanium-68 rod transmission sources. Such an acquisition protocol would increase the total scanning time beyond acceptable limits. The aim of this work is to validate the use of iterative reconstruction methods, on both transmission and emission scans, in order to obtain a fully corrected whole-body study within a reasonable scanning time of 60 min. Five minute emission and 3-min transmission scans are acquired at each of the seven bed positions. The transmission data are reconstructed with OSEM (ordered subsets expectation maximization) and the last iteration is reprojected to obtain consistent attenuation correction factors (ACFs). The emission image is then also reconstructed with OSEM, using the emission scan corrected for normalization, scatter and decay together with the set of consistent ACFs as inputs. The total processing time is about 35 min, which is acceptable in a clinical environment. The image quality, readability and accuracy of uptake quantification were assessed in 38 patients scanned for various malignancies. The sensitivity for tumour detection was the same for the non-attenuation-corrected (NAC-FBP) and the attenuation-corrected (AC-OSEM) images. The AC-OSEM images were less noisy and easier to interpret. The interobserver reproducibility was significantly increased when compared with non-corrected images (96.1% vs 81.1%, P<0.01). Standardized uptake values (SUVs) measured on images reconstructed with OSEM (AC-OSEM) and filtered backprojection (AC-FBP) were similar in all body regions except in the pelvic area, where SUVs were higher on AC-FBP images (mean increase 7.74%, P<0.01). Our results show that, when statistical reconstruction is applied to both transmission and emission data, high quality quantitative whole-body images are obtained within a reasonable scanning (60 min) and processing time, making it applicable in clinical practice. Received 27 November 1998 and in revised form 31 January 1999     

Effects of iterative reconstruction on image contrast and lesion detection in gamma camera coincidence imaging in lung and breast cancers.

To investigate the effects of iterative reconstruction in 18F-fluorodeoxyglucose (FDG) gamma camera coincidence imaging (GCI), image contrast and visual detection obtained by using the iterative ordered-subsets expectation maximization (OSEM) reconstruction, in a phantom and in patients with lung cancer and breast cancer, were compared with those obtained by using the conventional filtered backprojection (FBP) reconstruction. Images of a cylindrical phantom containing hot spheres of various sizes (10-38 mm) were acquired by positron emission tomography (PET) and GCI at various sphere-to-background activity ratios. Forty-one consecutive patients with biopsy-proven cancer of lung (n = 20) and breast (n = 21) underwent PET and GCI on the same day after intravenous injection of 370 MBq of FDG. GCI images reconstructed by the OSEM and the FBP were compared. FDG PET was considered as the standard of reference. In GCI phantom images, OSEM yielded better contrast and signal-to-noise ratio (SNR) than FBP over the range of sphere sizes. Attenuation correction improved both the image measures and sphere detection obtained by the OSEM in GCI. In the study involving patients, FDG PET depicted 41 primary tumours and 25 metastatic lymph nodes. All of the tumours >2 cm in diameter (n = 25), six of the nine tumours 1.5-2.0 cm in diameter (67%), two of seven tumours <1.5 cm in diameter (29%), and 20 metastatic lymph nodes (80%) were detected in attenuation uncorrected GCI reconstructed by the OSEM as well as the FBP. The undetected lesions in GCI were identical between the OSEM and the FBP reconstructions. OSEM yielded significantly greater tumour-to-background (T/B) ratios and lower noise than FBP in GCI (T/B ratios, 4.1+/-3.2 vs 3.7+/-2.7, P = 0.02; noise, 0.09+/-0.04 vs 0.14+/-0.05, P<0.0001). In conclusion, OSEM yielded better image contrast and less noise than the FBP in GCI, but the lesion detection obtained by the OSEM and the FBP in attenuation uncorrected GCI in patients with lung cancer and breast cancer were similar. Phantom data suggest the potential of OSEM for improving lesion detection in GCI after attenuation correction.     

Human-observer receiver-operating-characteristic evaluation of attenuation, scatter, and resolution compensation strategies for (99m)Tc myocardial perfusion imaging.

Nonuniform attenuation, scatter, and distance-dependent resolution are confounding factors inherent in SPECT imaging. Iterative reconstruction algorithms permit modeling and compensation of these degradations. We investigated through human-observer receiver-operating-characteristic (ROC) studies which (if any) combination of such compensation strategies best improves the accuracy of detection of coronary artery disease (CAD) when expert readers have only stress images for diagnosis. METHODS: A 3-headed SPECT system fitted with a (153)Gd line source was used to acquire simultaneously (99m)Tc-methoxyisobutylisonitrile (MIBI) images and transmission data. With these acquisitions, the accuracy of detecting CAD was evaluated for the following reconstruction strategies: filtered backprojection (FBP); ordered-subset expectation maximization (OSEM) with attenuation correction (AC); OSEM with AC and scatter correction (SC) (AC + SC); and OSEM with AC, SC, and resolution compensation (RC) (AC + SC + RC). Reconstruction parameters for OSEM were optimized by use of human-observer ROC studies with hybrid images, whereas standard clinical parameters were used for FBP. A total of 100 patients, including 55 patients referred for angiography and 45 patients with <5% likelihood for CAD, were included in the ROC studies. Images reconstructed with the 4 methods were rated independently with regard to the presence of CAD by 7 observers using a continuous scale for certainty. RESULTS: With area under the ROC curve (A(z)) as the criterion, the iterative reconstructions with compensation strategies (AC, AC + SC, and AC + SC + RC) demonstrated better detection accuracy than did FBP reconstructions for the overall detection of CAD as well as for the localization of perfusion defects in the 3 vascular territories. In general, the trend was for an increase in the A(z) for the progression from FBP to OSEM with AC, to OSEM with AC + SC, and to OSEM with AC + SC + RC. Statistically, the combination strategy with AC + SC + RC provided significantly higher A(z) values than did FBP images for the overall detection of CAD and the localization of perfusion defects in the left anterior descending coronary artery and left circumflex coronary artery territories, whereas AC + SC provided significantly better performance in the right coronary artery territory. CONCLUSION: The results indicate that OSEM with AC + SC + RC outperforms FBP reconstructions, indicating that the modeling of physical degradations can improve the accuracy of detection of CAD with cardiac perfusion SPECT reconstructions.     

A new iterative reconstruction technique for attenuation correction in high-resolution positron emission tomography

A new iterative reconstruction technique (NIRT) for positron emission computed tomography (PET), which uses transmission data for nonuniform attenuation correction, is described. Utilizing the general inverse problem theory, a cost functional which includes a noise term was derived. The cost functional was minimized using a weighted-least-square maximum a posteriori conjugate gradient (CG) method. The procedure involves a change in the Hessian of the cost function by adding an additional term. Two phantoms were used in a real data acquisition. The first was a cylinder phantom filled with uniformly distributed activity of 74 MBq of fluorine-18. Two different inserts were placed in the phantom. The second was a Hoffman brain phantom filled with uniformly distributed activity of 7.4 MBq of¹⁸F. Resulting reconstructed images were used to test and compare a new iterative reconstruction technique with a standard filtered backprojection (FBP) method. The results confirmed that NIRT, based on the conjugate gradient method, converges rapidly and provides good reconstructed images. In comparison with standard results obtained by the FBP method, the images reconstructed by NIRT showed better noise properties. The noise was measured as rms% noise and was less, by a factor of 1.75, in images reconstructed by NIRT than in the same images reconstructed by FBP. The distance between the Hoffman brain slice reconstructed by FBP and the perfect PET Hoffman brain slice created from the MRI image was 0.526, while the same distance for the Hoffman brain slice reconstructed by NIRT was 0.328. The NIRT method suppressed the propagation of the noise without visible loss of resolution in the reconstructed PET images.     

Improved tolerance to missing data in myocardial perfusion SPET using OSEM reconstruction

When projection data are incomplete for various technical reasons, artefacts may occur in the reconstructed images. This study examines whether an iterative reconstruction method, the ordered subsets implementation of the EM algorithm (OSEM), can improve reconstruction and minimise the artefacts compared to filtered back-projection (FBP). We varied the number and location of projections removed to investigate when significant artefacts occur, and whether diagnosis is affected. Phantom studies were analysed with sequential orthogonal pairs of projection angles removed (as would typically occur when either data loss or severe motion is detected during acquisition with a right-angled, dual-head cardiac single-photon emission tomography system) and reconstructed with both FBP and OSEM. Twelve normal myocardial perfusion studies were also assessed to study the effect of missing projections on clinical diagnosis. Differences between reconstructions with intact versus missing data were measured. Also, reconstructed images were clinically assessed and scored on a five-point scale based on whether the artefacts would alter clinical interpretation. Although both reconstruction methods showed artefacts, the absolute differences between reconstructed phantom data with intact and missing projection sets were significantly greater (P<0.005) for FBP than for OSEM for all numbers of missing projections. The clinical data showed similar differences between FBP and OSEM reconstructions. The three observers noted superiority of OSEM compared to FBP, with reduced incidence of clinically significant artefacts. However, neither reconstruction method could tolerate six or more missing pairs from 32 projections. There was no significant dependence on the angular location of missing projections. In the absence of any attempt to correct for missing projections, OSEM reduced the influence of artefacts compared to FBP.     

A comparison of the uniformity requirements for SPECT image reconstruction using FBP and OSEM techniques

OBJECTIVE: Gamma camera nonuniformity can result in the presence of ring artifacts in reconstructed SPECT images. The objective of this study is to compare the relationship between ring artifact magnitude and image noise in tomographic images reconstructed using FBP and OSEM. METHODS: A cylindrical phantom was filled with water and (99m)TC: Seven tomographic acquisitions were performed, with total counts per acquisition ranging from 1.5 Mcts to 100 MCTS: All acquisitions were reconstructed using both FBP and OSEM. Ring artifacts were generated in the transaxial data by introducing defects at a given location in each planar image. The modified acquisitions were again reconstructed using both FBP and OSEM. The ring artifacts were isolated by the subtraction of the uncorrupted datasets from the corrupted datasets. The magnitude of the ring artifacts in the corrupted reconstructions was measured and compared to the mean counts and noise level in the uncorrupted data. RESULTS: Ring magnitude in OSEM-reconstructed images is approximately one third that of FBP images. However, there is a corresponding reduction in image noise with OSEM and the ratio of ring magnitude-to-image noise was relatively similar for both OSEM and FBP. Rings generated with OSEM fell off more rapidly with distance from the image center, and reached a plateau at a higher magnitude at large distances. The visibility of rings with OSEM relative to FBP will depend on the location of the causative defect in the planar data and the number of iterations performed with OSEM. Differences between the 2 algorithms are subtle. CONCLUSION: Our results would indicate that the uniformity requirements for SPECT are similar for FBP and OSEM reconstruction algorithms.     

Effects of image reconstruction algorithm on neurotransmission PET studies in humans: Comparison between filtered backprojection and ordered subsets expectation maximization

OBJECTIVES: Both reconstruction algorithms, filtered backprojection (FBP) and ordered subsets expectation maximization (OSEM), are widely used in clinical positron emission tomography (PET) studies. Image reconstruction for most neurotransmission PET scan data is performed by FBP, while image reconstruction for whole-body [18F]FDG scan data is usually performed by OSEM. Although several investigators have compared FBP and OSEM in terms of the quantification of regional radioactivity and physiological parameters calculated from PET data, only a few studies have compared the two reconstruction algorithms in PET studies that estimate neurotransmission, i.e., neuroreceptor and neurotransporter binding. In this study we compared mean regional radioactivity concentration in the late phase and binding potential (BP) between FBP and OSEM algorithms in neurotransmission PET studies for [11C]raclopride and [11C]DASB. METHODS: Dynamic PET scans with [11C]raclopride in 3-dimensional mode were performed on seven healthy subjects. Dynamic PET scans with [11C]DASB in 2-dimensional mode were performed on another seven subjects. OSEM images were post-filtered so that its transverse spatial resolution became similar to that of FBP with the same Hanning filter (Kernel FWHM 6 mm). In both PET studies we calculated the BP of [11C]raclopride and [11C]DASB by a reference tissue model for each ROI (region of interest). RESULTS: There was no significant difference in mean regional radioactivity concentration between FBP and OSEM for [11C]raclopride and [11C]DASB. Only +2.4 - +3.2%, but still a significant difference in BP of [11C]raclopride between FBP and OSEM was observed in the striatum. There was no significant difference in BP between FBP and OSEM in other than the striatum for [11C]raclopride and in all regions for [11C]DASB. In addition, there was no significant difference in root mean square error between FBP and OSEM when BP was calculated. CONCLUSIONS: The BP values were similar between FBP and OSEM algorithms with [11C]raclopride and [11C]DASB. This study indicates that OSEM can be used for human neurotransmission PET studies for calculating BP although OSEM was not necessarily superior to FBP in the present study.     

Comparison of image reconstruction algorithms in myocardial perfusion scintigraphy

The purpose of this study was to compare the clinical utility of two image reconstruction algorithms in myocardial perfusion SPECT (single-photon emission computed tomography): filtered back-projection (FBP) and ordered subset expectation maximization (OSEM). A rest/stress one-day protocol with 99mTc-MIBI or tetrofosmin was performed on 102 consecutive patients who underwent coronary angiography. After SPECT data acquisition, images were reconstructed with FBP and OSEM algorithms. We assessed diagnostic performance (sensitivity, specificity and accuracy) in detecting coronary artery stenosis and evaluated regional tracer uptake with a 4-point scoring system. Although there were no significant differences in diagnostic performance between FBP and OSEM reconstruction, the OSEM method yielded higher uptake in the RCA area than the FBP method by reducing the count-loss artifact due to hepatic uptake of the tracers. In addition, regional uptake in the LCX area was significantly lower in the OSEM image than in the FBP image; this phenomenon was observed mainly in patients with coronary stenosis and/or infarction in the LCX territory. In conclusion, OSEM and FBP offered comparable diagnostic performance in stress myocardial perfusion SPECT. The OSEM method contributed to reduction of the count-loss artifact in inferior and posterior walls and to easy recognition of hypoperfusion in the LCX area.     

Evaluation of reconstruction techniques for lung single photon emission tomography: a Monte Carlo study

BACKGROUND: In studies of the distribution of lung function, the image quality of lung single photon emission computed tomography (SPECT) is important and one factor influencing it is the reconstruction algorithm. AIM: To systematically evaluate ordered subsets expectation maximization (OSEM) and compare it with filtered back-projection (FBP) for lung SPECT with Tc. METHODS: The evaluation of the number of iterations used in OSEM was based on the image quality parameter contrast. The comparison between OSEM and FBP was based on trade-off plots between statistical noise and spatial resolution for different filter parameters, collimators and count-levels. A Monte Carlo technique was used to simulate SPECT studies of a digital thorax phantom containing two sets of activity: one with a homogeneous activity distribution within the lungs and the other with superposed high- and low-activity objects. Statistical noise in the reconstructed images was calculated as the coefficient of variation (CV) and spatial resolution as full width at half-maximum (FWHM). RESULTS: For the configuration studied, the OSEM reconstruction in combination with post-filtering should be used in lung SPECT studies with at least 60 MLEM equivalent iterations. Compared to FBP the spatial resolution was improved by about 1 mm. For a constant level of CV, a four-fold increase in count level resulted in an increased resolution of about 2 mm. Spatial resolution and cut-off frequency depends on what value of noise in the image is acceptable also increased by using a low-energy, high-resolution collimator for CV values above 3%. The choice of noise-reducing filter and cut-off frequency depends on what value of noise in the image is acceptable.     

Evaluation of reconstruction techniques in regional cerebral blood flow SPECT using trade-off plots: a Monte Carlo study

BACKGROUND AND AIM: The image quality of single photon emission computed tomography (SPECT) depends on the reconstruction algorithm used. The purpose of the present study was to evaluate parameters in ordered subset expectation maximization (OSEM) and to compare systematically with filtered back-projection (FBP) for reconstruction of regional cerebral blood flow (rCBF) SPECT, incorporating attenuation and scatter correction. METHODS: The evaluation was based on the trade-off between contrast recovery and statistical noise using different sizes of subsets, number of iterations and filter parameters. Monte Carlo simulated SPECT studies of a digital human brain phantom were used. The contrast recovery was calculated as measured contrast divided by true contrast. Statistical noise in the reconstructed images was calculated as the coefficient of variation in pixel values. RESULTS: A constant contrast level was reached above 195 equivalent maximum likelihood expectation maximization iterations. The choice of subset size was not crucial as long as there were > or = 2 projections per subset. The OSEM reconstruction was found to give 5-14% higher contrast recovery than FBP for all clinically relevant noise levels in rCBF SPECT. The Butterworth filter, power 6, achieved the highest stable contrast recovery level at all clinically relevant noise levels. The cut-off frequency should be chosen according to the noise level accepted in the image. CONCLUSION: Trade-off plots are shown to be a practical way of deciding the number of iterations and subset size for the OSEM reconstruction and can be used for other examination types in nuclear medicine.