arXiv subjects

Hossein Arabi

Publications and source records attributed to Hossein Arabi.

28 records · Page 2Linked to original sources

arXiv2021

Comparison of different deep learning architectures for synthetic CT generation from MR images

MRI-guided radiation treatment planning is widely applied because of its superior soft-tissue contrast and no ionization radiation compared to CT-based planning. In this regard, synthetic CT (sCT) images should be generated from the patients MRI scans if radiation treatment planning is sought. Among the different available methods for this purpose, the deep learning algorithms have and do outperform their conventional counterparts. In this study, we investigated the performance of some most popular deep learning architectures including eCNN, U-Net, GAN, V-Net, and Res-Net for the task of sCT generation. As a baseline, an atlas-based method is implemented to which the results of the deep learning-based model are compared. A dataset consisting of 20 co-registered MR-CT pairs of the male pelvis is applied to assess the different sCT production methods' performance. The mean error (ME), mean absolute error (MAE), Pearson correlation coefficient (PCC), structural similarity index (SSIM), and peak signal-to-noise ratio (PSNR) metrics were computed between the estimated sCT and the ground truth (reference) CT images. The visual inspection revealed that the sCTs produced by eCNN, V-Net, and ResNet, unlike the other methods, were less noisy and greatly resemble the ground truth CT image. In the whole pelvis region, the eCNN yielded the lowest MAE (26.03-8.85 HU) and ME (0.82-7.06 HU), and the highest PCC metrics were yielded by the eCNN (0.93-0.05) and ResNet (0.91-0.02) methods. The ResNet model had the highest PSNR of 29.38-1.75 among all models. In terms of the Dice similarity coefficient, the eCNN method revealed superior performance in major tissue identification (air, bone, and soft tissue). All in all, the eCNN and ResNet deep learning methods revealed acceptable performance with clinically tolerable quantification errors.

physics.med-ph↗

arXiv2021

Automated lung segmentation from CT images of normal and COVID-19 pneumonia patients

Automated semantic image segmentation is an essential step in quantitative image analysis and disease diagnosis. This study investigates the performance of a deep learning-based model for lung segmentation from CT images for normal and COVID-19 patients. Chest CT images and corresponding lung masks of 1200 confirmed COVID-19 cases were used for training a residual neural network. The reference lung masks were generated through semi-automated/manual segmentation of the CT images. The performance of the model was evaluated on two distinct external test datasets including 120 normal and COVID-19 subjects, and the results of these groups were compared to each other. Different evaluation metrics such as dice coefficient (DSC), mean absolute error (MAE), relative mean HU difference, and relative volume difference were calculated to assess the accuracy of the predicted lung masks. The proposed deep learning method achieved DSC of 0.980 and 0.971 for normal and COVID-19 subjects, respectively, demonstrating significant overlap between predicted and reference lung masks. Moreover, MAEs of 0.037 HU and 0.061 HU, relative mean HU difference of -2.679% and -4.403%, and relative volume difference of 2.405% and 5.928% were obtained for normal and COVID-19 subjects, respectively. The comparable performance in lung segmentation of the normal and COVID-19 patients indicates the accuracy of the model for the identification of the lung tissue in the presence of the COVID-19 induced infections (though slightly better performance was observed for normal patients). The promising results achieved by the proposed deep learning-based model demonstrated its reliability in COVID-19 lung segmentation. This prerequisite step would lead to a more efficient and robust pneumonia lesion analysis.

eess.IV↗

arXiv2021

Deep learning-based noise reduction in low dose SPECT Myocardial Perfusion Imaging: Quantitative assessment and clinical performance

Clinical SPECT-MPI images of 345 patients acquired from a dedicated cardiac SPECT in list-mode format were retrospectively employed to predict normal-dose images from low-dose data at the half, quarter, and one-eighth-dose levels. A generative adversarial network was implemented to predict non-gated normal-dose images in the projection space at the different reduced dose levels. Established metrics including the peak signal-to-noise ratio (PSNR), root mean squared error (RMSE), and structural similarity index metrics (SSIM) in addition to Pearson correlation coefficient analysis and derived parameters from Cedars-Sinai software were used to quantitatively assess the quality of the predicted normal-dose images. For clinical evaluation, the quality of the predicted normal-dose images was evaluated by a nuclear medicine specialist using a seven-point (-3 to +3) grading scheme. By considering PSNR, SSIM, and RMSE quantitative parameters among the different reduced dose levels, the highest PSNR (42.49) and SSIM (0.99), and the lowest RMSE (1.99) were obtained at the half-dose level in the reconstructed images. Pearson correlation coefficients were measured 0.997, 0.994, and 0.987 for the predicted normal-dose images at the half, quarter, and one-eighth-dose levels, respectively. Regarding the normal-dose images as the reference, the Bland-Altman plots sketched for the Cedars-Sinai selected parameters exhibited remarkably less bias and variance in the predicted normal-dose images compared with the low-dose data at the entire reduced dose levels. Overall, considering the clinical assessment performed by a nuclear medicine specialist, 100%, 80%, and 11% of the predicted normal-dose images were clinically acceptable at the half, quarter, and one-eighth-dose levels, respectively.

physics.med-ph↗

arXiv2021

Quantitative investigation of low-dose PET imaging and post-reconstruction smoothing

In this study, 18F-FDG PET/CT brain scans of 50 patients with head and neck malignant lesions were employed to systematically assess the relationship between the amount of injected dose (10%, 8%, 6%, 5%, 4%, 3%, 2%, and 1% of standard dose) and the image quality through measuring standard image quality metrics (peak-signal-to-noise-ration (PSNR), structural similarity index (SSIM), root mean square error (RMSE), and standard uptake value (SUV) bias) for the whole head region as well as within the malignant lesions, considering the standard-dose PET images as reference. Furthermore, we evaluated the impact of post-reconstruction Gaussian filtering on the PET images in order to reduce noise and improve the signal-to-noise ratio at different low-dose levels. Significant degradation of PET image quality and tumor detectability was observed with a decrease in the injected dose by more than 5%, leading to a remarkable increase in RMSE from 0.173 SUV (at 5%) to 1.454 SUV (at 1%). The quantitative investigation of the malignant lesions demonstrated that SUVmax bias greatly increased in low-dose PET images (in particular at 1%, 2%, 3% levels) before applying the post-reconstruction filter, while applying the Gaussian filter on low-dose PET images led to a significant reduction in SUVmax bias. The SUVmean bias within the malignant lesions was negligible (less than 1%) in low-dose PET images; however, this bias increased significantly after applying the post-reconstruction filter. In conclusion, it is strongly recommended that the SUVmax bias and SUVmean bias in low-dose PET images should be considered prior to and following the application of the post-reconstruction Guassain filter, respectively.

physics.med-ph↗

arXiv2021

Deep learning-based synthetic CT generation from MR images: comparison of generative adversarial and residual neural networks

Currently, MRI-only radiotherapy (RT) eliminates some of the concerns about using CT images in RT chains such as the registration of MR images to a separate CT, extra dose delivery, and the additional cost of repeated imaging. However, one remaining challenge is that the signal intensities of MRI are not related to the attenuation coefficient of the biological tissue. This work compares the performance of two state-of-the-art deep learning models; a generative adversarial network (GAN) and a residual network (ResNet) for synthetic CTs (sCT) generation from MR images. The brain MR and CT images of 86 participants were analyzed. GAN and ResNet models were implemented for the generation of synthetic CTs from the 3D T1-weighted MR images using a six-fold cross-validation scheme. The resulting sCTs were compared, considering the CT images as a reference using standard metrics such as the mean absolute error (MAE), peak signal-to-noise ratio (PSNR) and the structural similarity index (SSIM). Overall, the ResNet model exhibited higher accuracy in relation to the delineation of brain tissues. The ResNet model estimated the CT values for the entire head region with an MAE of 114.1 HU compared to MAE=-10.9 HU obtained from the GAN model. Moreover, both models offered comparable SSIM and PSNR values, although the ResNet method exhibited a slightly superior performance over the GAN method. We compared two state-of-the-art deep learning models for the task of MR-based sCT generation. The ResNet model exhibited superior results, thus demonstrating its potential to be used for the challenge of synthetic CT generation in PET/MR AC and MR-only RT planning.

physics.med-ph↗

arXiv2021

Monte Carlo-based estimation of patient absorbed dose in 99mTc-DMSA, -MAG3, and -DTPA SPECT imaging using the University of Florida (UF) phantoms

SPECT imaging is one of the common modalities used to examine kidney disease. The radiopharmaceuticals such as 99mTc-MAG3, 99mTc-DTPA, and 99mTc-DMSA are commonly employed for pediatric patients. It is generally accepted that children organs are more sensitive to radiation than adults, due to their growth rate. Therefore, evaluating the absorbed dose in children to avoid irrecoverable damage is highly crucial. In this work, absorbed dose by different organs of children within the SPECT imaging for the above-mentioned radiopharmaceuticals were estimated through the use of Monte Carlo simulation for patients at the ages of 4, 8, 11, and 14-years old. The patients absorbed dose was estimated based on the specific radiopharmaceutical uptake patterns and the age-related human body models. To this end, we used the University of Florida (UF) voxel-wise baby phantoms and the Monte Carlo simulation for the pediatric patients at 4, 8, 11, and 14 years of age. The accuracy of the dosimetric outcomes was evaluated versus the ICRP 128 data. The highest and lowest doses absorbed by the kidneys were 0.48 mGy-MBq and 0.0042 mGy-MBq obtained for 4-year-old and 14-year-old children using 99mTc-DMSA and 99mTc-MAG3 radiopharmaceuticals, respectively. Moreover, in these cases, the relative errors of simulation results compared to the reference ICRP data were 12.1% and 0.95%, respectively. The simulation results were in good agreement with the ICRP 128 data. The results showed that the highest absorbed dose was by kidneys and when 99mTc-DMSA was used. The highest and lowest absorbed dose in the organs occurs when 99mTc-DMSA and 99mTc-MAG3 (in normal renal function) are used, respectively. On the other hand, the results indicated that the dose received by the organs decreases with age.

physics.med-ph↗

arXiv2021

Quantitative analysis of image quality in low-dose CT imaging for Covid-19 patients

We set out to simulate four reduced dose-levels (60%-dose, 40%-dose, 20%-dose, and 10%-dose) of standard CT imaging using Beer-Lambert's law across 49 patients infected with COVID-19. Then, three denoising filters, namely Gaussian, Bilateral, and Median, were applied to the different low-dose CT images, the quality of which was assessed prior to and after the application of the various filters via calculation of peak signal-to-noise ratio (PSNR), root mean square error (RMSE), structural similarity index measure (SSIM), and relative CT-value bias, separately for the lung tissue and whole-body. The quantitative evaluation indicated that 10%-dose CT images have inferior quality (with RMSE=322.1-+104.0 HU and bias=11.44-+4.49% in the lung) even after the application of the denoising filters. The bilateral filter exhibited superior performance to suppress the noise and recover the underlying signals in low-dose CT images compared to the other denoising techniques. The bilateral filter led to RMSE and bias of 100.21-+16.47 HU, -0.21-+1.20%, respectively in the lung regions for 20%-dose CT images compared to the Gaussian filter with RMSE=103.46-+15.70 HU and bias=1.02-+1.68%, median filter with RMSE=129.60-+18.09 HU and bias=-6.15-+2.24%, and the nonfiltered 20%-dose CT with RMSE=217.37-+64.66 HU and bias=4.30-+1.85%. In conclusion, the 20%-dose CT imaging followed by the bilateral filtering introduced a reasonable compromise between image quality and patient dose reduction.

physics.med-ph↗

arXiv2021

Dosimetric characterization of a new 192Ir pulse dose rate brachytherapy source with the Monte Carlo simulation and thermoluminescent dosimeter

In this study, recommendations of the AAPM TG- 43 (U1) report have been followed to characterize the new 192Ir pulse dose rate source, provided by the Applied Radiation Research School, Nuclear Science and Technology Research Institute in Iran. Dose rate constant, radial dose function, geometry factors, and anisotropy function were calculated according to the relevant American Association of Physicists in Medicine AAPM and TG43 (U1) reports. In this study, 192Ir source was characterized using Monte Carlo simulation in water phantom, and in addition, experimental measurements were carried out using thermoluminescent dosimeters (TLD-100) in plexiglass (PMMA) phantoms. The dose-rate constant for the 192Ir PDR was found to be equal to 1.131 cGyh-1U-1 and 1.173 cGyh-1U-1 with TLD measurement and Monte Carlo simulation, respectively. Also in this study, the geometry function, radial dose functions g(r), and the anisotropy function have been calculated at distances from 0.1 to 16 cm. The dose-rate constant of these calculations has been compared with measured values for an actual 192Ir seed. The results of dosimetry parameters, presented in tabulated and graphical formats, exhibited good agreement to those reported from other commercially available PDR 192Ir sources. The results obtained in this study are in close agreement with the characteristics of the commercially available 192Ir sources. The results obtained in this study can be treated as an initial assessment of this source to be employed in the conventional treatment planning systems subsequent to complementary investigations.

physics.med-ph↗

arXiv2021

Deep learning-based attenuation correction in the image domain for myocardial perfusion SPECT imaging

Objective: In this work, we set out to investigate the accuracy of direct attenuation correction (AC) in the image domain for the myocardial perfusion SPECT imaging (MPI-SPECT) using two residual (ResNet) and UNet deep convolutional neural networks. Methods: The MPI-SPECT 99mTc-sestamibi images of 99 participants were retrospectively examined. UNet and ResNet networks were trained using SPECT non-attenuation corrected images as input and CT-based attenuation corrected SPECT images (CT-AC) as reference. The Chang AC approach, considering a uniform attenuation coefficient within the body contour, was also implemented. Quantitative and clinical evaluation of the proposed methods were performed considering SPECT CT-AC images of 19 subjects as reference using the mean absolute error (MAE), structural similarity index (SSIM) metrics, as well as relevant clinical indices such as perfusion deficit (TPD). Results: Overall, the deep learning solution exhibited good agreement with the CT-based AC, noticeably outperforming the Chang method. The ResNet and UNet models resulted in the ME (count) of ${-6.99\pm16.72}$ and ${-4.41\pm11.8}$ and SSIM of ${0.99\pm0.04}$ and ${0.98\pm0.05}$, respectively. While the Change approach led to ME and SSIM of ${25.52\pm33.98}$ and ${0.93\pm0.09}$, respectively. Similarly, the clinical evaluation revealed a mean TPD of ${12.78\pm9.22}$ and ${12.57\pm8.93}$ for the ResNet and UNet models, respectively, compared to ${12.84\pm8.63}$ obtained from the reference SPECT CT-AC images. On the other hand, the Chang approach led to a mean TPD of ${16.68\pm11.24}$. Conclusion: We evaluated two deep convolutional neural networks to estimate SPECT-AC images directly from the non-attenuation corrected images. The deep learning solutions exhibited the promising potential to generate reliable attenuation corrected SPECT images without the use of transmission scanning.

physics.med-ph↗

arXiv2021

Comparison of the x-ray tube spectrum measurement using BGO, NaI, LYSO, and HPGe detectors in a preclinical mini-CT scanner: Monte Carlo simulation and practical experiment

The aim of this study is to estimate the intrinsic efficiency and energy resolution of different types of solid-state gamma-ray detectors in order to generate a precise dual-energy x-ray beam from the conventional x-ray tube using external x-ray filters. The x-ray spectrum of a clinical x-ray tube was experimentally measured using a high purity Germanium detector (HPGe) and the obtained spectrum validated by Monte Carlo (MC) simulations. The obtained x-ray spectrum from the experiment was employed to assess the energy resolution and detection efficiency of different inorganic scintillators and semiconductor-based solid-state detectors, namely HPGe, BGO, NaI, and LYSO, using MC simulations. The best performing detector was employed to experimentally create and measure a dual-energy x-ray spectrum through applying attenuating filters to the original x-ray beam. The simulation results indicated a 9.16% energy resolution for the HPGe detector wherein the FWHM of the energy resolution for the HPGe detector was about 1/3rd of the other inorganic detectors. The x-ray spectra estimated from the various source energies exhibited a good agreement between experimental and simulation results with a maximum difference of 6%. Owing to the high-energy discrimination power of the HPGe detector, a dual-energy x-ray spectrum was created and measured from the original x-ray spectrum using 0.5 and 4.5 mm Aluminum external filters, which involves 70 and 140 keV energy peaks with 8% overlap. The experimental measurements and MC simulations of the HPGe detector exhibited close agreement in high-energy resolution estimation of the x-ray spectrum. Given the accurate measurement of the x-ray spectrum with the HPGe detector, a dual-energy x-ray spectrum was generated with minimal energy overlap using external x-ray filters.

physics.ins-det↗