Objective： To explore whether single and fused monochromatic images can improve liver tumor detection and delineation by single source dual energy CT （ssDECT） in patients with hepatocellular carcinoma （HCC） durin...Objective： To explore whether single and fused monochromatic images can improve liver tumor detection and delineation by single source dual energy CT （ssDECT） in patients with hepatocellular carcinoma （HCC） during arterial phase. Methods： Fifty-seven patients with HCC who underwent ssDECT scanning at Beijing Cancer Hospital were enrolled retrospectively. Twenty-one sets of monochromatic images from 40 to 140 keV were reconstructed at 5 keV intervals in arterial phase. The optimal contrast-noise ratio （CNR） monochromatic images of the liver tumor and the lowest-noise monochromatic images were selected for image fusion. We evaluated the image quality of the optimal-CNR monochromatic images, the lowest-noise monochromatic images and the fused monochromatic images, respectively. The evaluation indicators included the spatial resolution of the anatomical structure, the noise level, the contrast and CNR of the tumor. Results： In arterial phase, the anatomical structure of the liver can be displayed most clearly in the 65-keV monochromatic images, with the lowest image noise. The optimal-CNR monochromatic images of HCC tumor were 50-keV monochromatic images in which the internal structural features of the liver tumors were displayed most clearly and meticulously. For tumor detection, the fused monochromatic images and the 50-keV monochromatic images had similar performances, and were more sensitive than 65-keV monochromatic images. Conclusions： We achieved good arterial phase images by fusing the optimal-CNR monochromatic images of the HCC tumor and the lowest-noise monochromatic images. The fused images displayed liver tumors and anatomical structures more clearly, which is potentially helpful for identifying more and smaller HCC tumors.展开更多
This study proposes a novel feature extraction approach for radionuclide identification to increase the precision of identification of the gamma-ray energy spectrum set.For easier utilization of the information contai...This study proposes a novel feature extraction approach for radionuclide identification to increase the precision of identification of the gamma-ray energy spectrum set.For easier utilization of the information contained in the spectra,the vectors of the gamma-ray energy spectra from Euclidean space,which are fingerprints of the different types of radionuclides,were mapped to matrices in the Banach space.Subsequently,to make the spectra in matrix form easier to apply to image-based deep learning frameworks,the matrices of the gamma-ray energy spectra were mapped to images in the RGB color space.A deep convolutional neural network(DCNN)model was constructed and trained on the ImageNet dataset.The mapped gamma-ray energy spectrum images were applied as inputs to the DCNN model,and the corresponding outputs of the convolution layers and fully connected layers were transferred as descriptors of the images to construct a new classification model for radionuclide identification.The transferred image descriptors consist of global and local features,where the activation vectors of fully connected layers are global features,and activations from convolution layers are local features.A series of comparative experiments between the transferred image descriptors,peak information,features extracted by the histogram of the oriented gradients(HOG),and scale-invariant feature transform(SIFT)using both synthetic and measured data were applied to 11 classical classifiers.The results demonstrate that although the gamma-ray energy spectrum images are completely unfamiliar to the DCNN model and have not been used in the pre-training process,the transferred image descriptors achieved good classification results.The global features have strong semantic information,which achieves an average accuracy of 92.76%and 94.86%on the synthetic dataset and measured dataset,respectively.The results of the statistical comparison of features demonstrate that the proposed approach outperforms the peak-searching-based method,HOG,and SIFT on the synthetic and measured datasets.展开更多
The measurement of electron density is important for medical diagnosis and charged particle radiotherapy treatment planning.Traditionally,electron density is obtained by CT imaging using the relationship between CT-nu...The measurement of electron density is important for medical diagnosis and charged particle radiotherapy treatment planning.Traditionally,electron density is obtained by CT imaging using the relationship between CT-number and electron densities established beforehand.However,the measurement is not accurate due to the beam hardening effect.In this paper,we propose a simple and practical electron density acquisition method based on dual-energy CT technique.For each sample,the CT imaging is conducted using two selected X-ray energy from synchrotron radiation.A post-processing dual-energy reconstruction method is used.Linear attenuation coefficients of the scanned samples are obtained by FBP reconstruction.The effective atomic number and electron density are got by solving the dual-energy simultaneous equations.Different phantoms and breast tissues were scanned in this experimental study under 10 keV and 30 keV monochromatic X-rays.The distribution of effective atomic numbers and electron densities of the scanned phantoms were obtained by Dual-energy CT image reconstruction,which agrees well with the theoretical values.Compared with conventional methods,the measurement accuracy is greatly improved, and the measurement error is reduced to about 1%.This experimental study demonstrates that DECT imaging based on synchrotron radiation source is applicable to medical diagnosis for quantitative measurement with high accuracy.展开更多
Objective: Computed tomography (CT)-based attenuation correction (CTAC) offers the clear benefit of reliable reconstruction of single-photon emission computed tomography (SPECT) images through its ability to achieve o...Objective: Computed tomography (CT)-based attenuation correction (CTAC) offers the clear benefit of reliable reconstruction of single-photon emission computed tomography (SPECT) images through its ability to achieve object-specific attenuation maps, but artifacts from dense materials often deteriorate CTAC performance. Therefore, we investigate the feasibility of CTAC in the presence of dense materials using dual-energy virtual monochromatic CT data. Methods: A sodium pertechnetate-filled cylindrical uniform phantom, with a pair of undiluted iodine syringes attached, is scanned with a dual-source CT scanner to obtain both single-energy (120 kVp) polychromatic and dual-energy (80 kVp/140 kVp with tin filtering) virtual monochromatic CT images. The single-energy and the dual-energy CT images are then converted to attenuation maps at 141 keV. SPECT images are reconstructed from 99mTc emission data of the phantom using each single-energy and dual-energy attenuation map and incorporating CTAC procedure. A region-of-in- terest analysis is performed to quantitatively compare the attenuation maps between the single-energy and the dual-energy techniques, each at an iodine-free position and a position adjacent to the iodine solutions. Results: At the iodine-free position, the phantom provides a uniform distribution of attenuation maps in both the single-energy and the dual-energy techniques. In the presence of adjacent iodine solutions, however, severe artifacts appeare in the single-energy CT images. These artifacts make attenuation values fluctuate, resulting in erroneous pixel values in the CTAC SPECT images. In contrast, dual-energy CT strongly suppresses the artifacts and hence improves the uniformity of the attenuation maps and the resultant SPECT images. Conclusions: Dual-energy CT with virtual monochromatic reconstruction has the potential to substantially reduce artifacts arising from dense materials. It has the potential to improve the accuracy of attenuation maps and the resultant CTAC SPECT images.展开更多
基金supported by the National Basic Research Program of China (973 Program) (Grant No. 2011CB707705)National Natural Science Foundation of China (Grant No. 81371715+1 种基金 81201215)the Capital Characteristic Clinical Application Research (Grant No. Z121107001012115)
文摘Objective： To explore whether single and fused monochromatic images can improve liver tumor detection and delineation by single source dual energy CT （ssDECT） in patients with hepatocellular carcinoma （HCC） during arterial phase. Methods： Fifty-seven patients with HCC who underwent ssDECT scanning at Beijing Cancer Hospital were enrolled retrospectively. Twenty-one sets of monochromatic images from 40 to 140 keV were reconstructed at 5 keV intervals in arterial phase. The optimal contrast-noise ratio （CNR） monochromatic images of the liver tumor and the lowest-noise monochromatic images were selected for image fusion. We evaluated the image quality of the optimal-CNR monochromatic images, the lowest-noise monochromatic images and the fused monochromatic images, respectively. The evaluation indicators included the spatial resolution of the anatomical structure, the noise level, the contrast and CNR of the tumor. Results： In arterial phase, the anatomical structure of the liver can be displayed most clearly in the 65-keV monochromatic images, with the lowest image noise. The optimal-CNR monochromatic images of HCC tumor were 50-keV monochromatic images in which the internal structural features of the liver tumors were displayed most clearly and meticulously. For tumor detection, the fused monochromatic images and the 50-keV monochromatic images had similar performances, and were more sensitive than 65-keV monochromatic images. Conclusions： We achieved good arterial phase images by fusing the optimal-CNR monochromatic images of the HCC tumor and the lowest-noise monochromatic images. The fused images displayed liver tumors and anatomical structures more clearly, which is potentially helpful for identifying more and smaller HCC tumors.
基金supported by the National Defense Fundamental Research Project(No.JCKY2020404C004)Sichuan Science and Technology Program(No.22NSFSC0044).
文摘This study proposes a novel feature extraction approach for radionuclide identification to increase the precision of identification of the gamma-ray energy spectrum set.For easier utilization of the information contained in the spectra,the vectors of the gamma-ray energy spectra from Euclidean space,which are fingerprints of the different types of radionuclides,were mapped to matrices in the Banach space.Subsequently,to make the spectra in matrix form easier to apply to image-based deep learning frameworks,the matrices of the gamma-ray energy spectra were mapped to images in the RGB color space.A deep convolutional neural network(DCNN)model was constructed and trained on the ImageNet dataset.The mapped gamma-ray energy spectrum images were applied as inputs to the DCNN model,and the corresponding outputs of the convolution layers and fully connected layers were transferred as descriptors of the images to construct a new classification model for radionuclide identification.The transferred image descriptors consist of global and local features,where the activation vectors of fully connected layers are global features,and activations from convolution layers are local features.A series of comparative experiments between the transferred image descriptors,peak information,features extracted by the histogram of the oriented gradients(HOG),and scale-invariant feature transform(SIFT)using both synthetic and measured data were applied to 11 classical classifiers.The results demonstrate that although the gamma-ray energy spectrum images are completely unfamiliar to the DCNN model and have not been used in the pre-training process,the transferred image descriptors achieved good classification results.The global features have strong semantic information,which achieves an average accuracy of 92.76%and 94.86%on the synthetic dataset and measured dataset,respectively.The results of the statistical comparison of features demonstrate that the proposed approach outperforms the peak-searching-based method,HOG,and SIFT on the synthetic and measured datasets.
基金supported by National Key Technology R&D Program of the Ministry of Science and Technology(No.2012BA107B05)
文摘The measurement of electron density is important for medical diagnosis and charged particle radiotherapy treatment planning.Traditionally,electron density is obtained by CT imaging using the relationship between CT-number and electron densities established beforehand.However,the measurement is not accurate due to the beam hardening effect.In this paper,we propose a simple and practical electron density acquisition method based on dual-energy CT technique.For each sample,the CT imaging is conducted using two selected X-ray energy from synchrotron radiation.A post-processing dual-energy reconstruction method is used.Linear attenuation coefficients of the scanned samples are obtained by FBP reconstruction.The effective atomic number and electron density are got by solving the dual-energy simultaneous equations.Different phantoms and breast tissues were scanned in this experimental study under 10 keV and 30 keV monochromatic X-rays.The distribution of effective atomic numbers and electron densities of the scanned phantoms were obtained by Dual-energy CT image reconstruction,which agrees well with the theoretical values.Compared with conventional methods,the measurement accuracy is greatly improved, and the measurement error is reduced to about 1%.This experimental study demonstrates that DECT imaging based on synchrotron radiation source is applicable to medical diagnosis for quantitative measurement with high accuracy.
文摘Objective: Computed tomography (CT)-based attenuation correction (CTAC) offers the clear benefit of reliable reconstruction of single-photon emission computed tomography (SPECT) images through its ability to achieve object-specific attenuation maps, but artifacts from dense materials often deteriorate CTAC performance. Therefore, we investigate the feasibility of CTAC in the presence of dense materials using dual-energy virtual monochromatic CT data. Methods: A sodium pertechnetate-filled cylindrical uniform phantom, with a pair of undiluted iodine syringes attached, is scanned with a dual-source CT scanner to obtain both single-energy (120 kVp) polychromatic and dual-energy (80 kVp/140 kVp with tin filtering) virtual monochromatic CT images. The single-energy and the dual-energy CT images are then converted to attenuation maps at 141 keV. SPECT images are reconstructed from 99mTc emission data of the phantom using each single-energy and dual-energy attenuation map and incorporating CTAC procedure. A region-of-in- terest analysis is performed to quantitatively compare the attenuation maps between the single-energy and the dual-energy techniques, each at an iodine-free position and a position adjacent to the iodine solutions. Results: At the iodine-free position, the phantom provides a uniform distribution of attenuation maps in both the single-energy and the dual-energy techniques. In the presence of adjacent iodine solutions, however, severe artifacts appeare in the single-energy CT images. These artifacts make attenuation values fluctuate, resulting in erroneous pixel values in the CTAC SPECT images. In contrast, dual-energy CT strongly suppresses the artifacts and hence improves the uniformity of the attenuation maps and the resultant SPECT images. Conclusions: Dual-energy CT with virtual monochromatic reconstruction has the potential to substantially reduce artifacts arising from dense materials. It has the potential to improve the accuracy of attenuation maps and the resultant CTAC SPECT images.
