Ultra-limited-angle CT image reconstruction algorithm based on reweighting and edge-preserving

BACKGROUND: Ultra-limited-angle image reconstruction problem with a limited-angle scanning range less than or equal to π 2 is severely ill-posed. Due to the considerably large condition number of a linear system for image reconstruction, it is extremely challenging to generate a valid reconstructed image by traditional iterative reconstruction algorithms. OBJECTIVE: To develop and test a valid ultra-limited-angle CT image reconstruction algorithm. METHODS: We propose a new optimized reconstruction model and Reweighted Alternating Edge-preserving Diffusion and Smoothing algorithm in which a reweighted method of improving the condition number is incorporated into the idea of AEDS image reconstruction algorithm. The AEDS algorithm utilizes the property of image sparsity to improve partially the results. In experiments, the different algorithms (the Pre-Landweber, AEDS algorithms and our algorithm) are used to reconstruct the Shepp-Logan phantom from the simulated projection data with noises and the flat object with a large ratio between length and width from the real projection data. PSNR and SSIM are used as the quantitative indices to evaluate quality of reconstructed images. RESULTS: Experiment results showed that for simulated projection data, our algorithm improves PSNR and SSIM from 22.46db to 39.38db and from 0.71 to 0.96, respectively. For real projection data, our algorithm yields the highest PSNR and SSIM of 30.89db and 0.88, which obtains a valid reconstructed result. CONCLUSIONS: Our algorithm successfully combines the merits of several image processing and reconstruction algorithms. Thus, our new algorithm outperforms significantly other two algorithms and is valid for ultra-limited-angle CT image reconstruction.

Automated detection of diabetic retinopathy using custom convolutional neural network

Diabetic retinopathy is an eye deficiency that affects retina as a result of the patient having diabetes mellitus caused by high sugar levels, which may eventually lead to macular edema. The objective of this study is to design and compare several deep learning models that detect severity of diabetic retinopathy, determine risk of leading to macular edema, and segment different types of disease patterns using retina images. Indian Diabetic Retinopathy Image Dataset (IDRiD) dataset was used for disease grading and segmentation. Since images of the dataset have different brightness and contrast, we employed three techniques for generating processed images from the original images, which include brightness, color and, contrast (BCC) enhancing, color jitters (CJ), and contrast limited adaptive histogram equalization (CLAHE). After image preporcessing, we used pre-trained ResNet50, VGG16, and VGG19 models on these different preprocessed images both for determining the severity of the retinopathy and also the chances of macular edema. UNet was also applied to segment different types of diseases. To train and test these models, image dataset was divided into training, testing, and validation data at 70%, 20%, and 10% ratios, respectively. During model training, data augmentation method was also applied to increase the number of training images. Study results show that for detecting the severity of retinopathy and macular edema, ResNet50 showed the best accuracy using BCC and original images with an accuracy of 60.2% and 82.5%, respectively, on validation dataset. In segmenting different types of diseases, UNet yielded the highest testing accuracy of 65.22% and 91.09% for microaneurysms and hard exudates using BCC images, 84.83% for optic disc using CJ images, 59.35% and 89.69% for hemorrhages and soft exudates using CLAHE images, respectively. Thus, image preprocessing can play an important role to improve efficacy and performance of deep learning models.

Image quality enhancement of CT hepatic portal venography using dual energy blending with computer determined parameters

BACKGROUND: Previous studies have shown that using some post-processing methods, such as nonlinear-blending and linear blending techniques, has potential to improve dual-energy computed (DECT) image quality. OBJECTIVE: To improve DECT image quality of hepatic portal venography (CTPV) using a new non-linear blending method with computer-determined parameters, and to compare the results to additional linear and non-linear blending techniques. METHODS: DECT images of 60 patients who were clinically diagnosed with liver cirrhosis were selected and studied. Dual-energy scanning (80 kVp and Sn140 kVp) of CTPV was utilized in the portal venous phase through a dual-source CT scanner. For image processing, four protocols were utilized including linear blending with a weighing factor of 0.3 (protocol A) and 1.0 (protocol B), non-linear blending with fixed blending width of 200 HU and set blending center of 150HU (protocol C), and computer-based blending (protocol D). Several image quality indicators, including signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR) and contrast of hepatic portal vein and hepatic parenchyma, were evaluated using the paired-sample t-test. A 5-grade scale scoring system was also utilized for subjective analysis. RESULTS: SNR of protocols A-D were 9.1±2.1, 12.1±3.0, 11.6±2.8 and 14.4±3.2, respectively. CNR of protocols A-D were 4.6±1.3, 8.0±2.3, 7.0±2.0 and 9.8±2.4, respectively. The contrast of protocols A-D were 37.7±11.6, 91.9±21.0, 66.2±19.0 and 107.7±21.3, respectively. The differences between protocol D and other three protocols were significant (P < 0.01). In subjective evaluation, the modes of protocols A, B, C, and D were rated poor, good, generally acceptable, and excellent, respectively. CONCLUSION: The non-linear blending technique of protocol D with computer-determined blending parameters can help improve imaging quality of CTPV and contribute to a diagnosis of liver disease.

Optical and radiation shielding characteristics of tellurite glass doped with different rare-earth oxides

Shielding glass materials doped with heavy metal oxides show an improvement in the effectiveness of the materials used in radiation shielding. In this work, the photon shielding parameters of six tellurite glass systems doped with several metal oxides namely, 70TeO2-10P2O5- 10ZnO- 5.0PbF2- 0.0024Er2O3- 5.0X (where X represents different doped metail oxides namely, Nb2O5, TiO2, WO3, PbO, Bi2O3, and CdO) in a broad energy spectrum, ranging from 0.015 MeV to 15 MeV, were evaluated. The shielding parameters were calculated using the online software Phy-X/PSD. The highest linear and mass attenuation coefficients recorded were obtaibed from the samples containing bismuth oxide (Bi2O3), and the lowest half-value layer and mean free path were recorded among the other samples. Furthermore, the shielding effectiveness of tellurite glass systems was compared with commercial shielding materials (RS-369, RS-253 G18, chromite, ferrite, magnetite, and barite). The optical parameters viz, dispersion energy, single-oscillator energy, molar refraction, electronic polarizability, non-linear refractive indices, n2 , and third-order susceptibility were measured and reported at a different wavelength. Bi2O3 has a strong effect on enhancing the optical and shielding properties. The outcome of this study suggests the potential of using the proposed glass samples as radiation-shielding materials for a broad range of imaging and therapeutic applications.

High-quality quasi-parallel X-ray beam obtained by a parabolic monocapillary X-ray lens with a square beam stop

Parabolic monocapillary X-ray lens (PMXRL) is an ideal optical device for constraining the point divergent X-ray beams to quasi-parallel beams, but the overlap of direct X-rays and reflected X-rays through PMXRL deteriorates the outgoing beam divergence. Aiming to solve this problem, this study designs and tests a square-shaped lead occluder (SSLO) embedded in PMXRL to block the direct X-rays passing through the PMXRL. Python simulations are employed to determine the geometric parameters of the SSLO as well as the optimal position of the SSLO in the PMXRL according to our proposed model. The PMXRL with a conic parameter p of 0.000939 mm and a length L of 60.8 mm is manufactured and the SSLO with a size of 0.472 mm×0.472 mm×3.4 mm is embedded into it. An optical path system based on this PMXRL is built to measure the divergence of the outgoing X-ray beam. The experimental results show that the quasi-parallel X-ray beam reaches a divergence of 0.36 mrad in the range from 15–45 mm at the PMXRL outlet. This divergence is 10 times lower than the theoretical divergence without SSLO. Our work provides an alternative method for obtaining highly parallel X-ray beam and is beneficial to generate or facilitate new applications of monocapillary optics in X-ray technology.

Evaluation and comparison of a CdTe based photon counting detector with an energy integrating detector for X-ray phase sensitive imaging of breast cancer

PURPOSE: To compare imaging performance of a cadmium telluride (CdTe) based photon counting detector (PCD) with a CMOS based energy integrating detector (EID) for potential phase sensitive imaging of breast cancer. METHODS: A high energy inline phase sensitive imaging prototype consisting of a microfocus X-ray source with geometric magnification of 2 was employed. The pixel pitch of the PCD was 55μm, while 50μm for EID. The spatial resolution was quantitatively and qualitatively assessed through modulation transfer function (MTF) and bar pattern images. The edge enhancement visibility was assessed by measuring edge enhancement index (EEI) using the acrylic edge acquired images. A contrast detail (CD) phantom was utilized to compare detectability of simulated tumors, while an American College of Radiology (ACR) accredited phantom for mammography was used to compare detection of simulated calcification clusters. A custom-built phantom was employed to compare detection of fibrous structures. The PCD images were acquired at equal, and 30% less mean glandular dose (MGD) levels as of EID images. Observer studies along with contrast to noise ratio (CNR) and signal to noise ratio (SNR) analyses were performed for comparison of two detection systems. RESULTS: MTF curves and bar pattern images revealed an improvement of about 40% in the cutoff resolution with the PCD. The excellent spatial resolution offered by PCD system complemented superior detection of the diffraction fringes at boundaries of the acrylic edge and resulted in an EEI value of 3.64 as compared to 1.44 produced with EID image. At MGD levels (standard dose), observer studies along with CNR and SNR analyses revealed a substantial improvement of PCD acquired images in detection of simulated tumors, calcification clusters, and fibrous structures. At 30% less MGD, PCD images preserved image quality to yield equivalent (slightly better) detection as compared to the standard dose EID images. CONCLUSION: CdTe-based PCDs are technically feasible to image breast abnormalities (low/high contrast structures) at low radiation dose levels using the high energy inline phase sensitive imaging technique.

A coronary artery CTA segmentation approach based on deep learning

Presence of plaque and coronary artery stenosis are the main causes of coronary heart disease. Detection of plaque and coronary artery segmentation have become the first choice in detecting coronary artery disease. The purpose of this study is to investigate a new method for plaque detection and automatic segmentation and diagnosis of coronary arteries and to test its feasibility of applying to clinical medical image diagnosis. A multi-model fusion coronary CT angiography (CTA) vessel segmentation method is proposed based on deep learning. The method includes three network layer models namely, an original 3-dimensional full convolutional network (3D FCN) and two networks that embed the attention gating (AG) model in the original 3D FCN. Then, the prediction results of the three networks are merged by using the majority voting algorithm and thus the final prediction result of the networks is obtained. In the post-processing stage, the level set function is used to further iteratively optimize the results of network fusion prediction. The JI (Jaccard index) and DSC (Dice similarity coefficient) scores are calculated to evaluate accuracy of blood vessel segmentations. Applying to a CTA dataset of 20 patients, accuracy of coronary blood vessel segmentation using FCN, FCN-AG1, FCN-AG2 network and the fusion method are tested. The average values of JI and DSC of using the first three networks are (0.7962, 0.8843), (0.8154, 0.8966) and (0.8119, 0.8936), respectively. When using new fusion method, average JI and DSC of segmentation results increase to (0.8214, 0.9005), which are better than the best result of using FCN, FCN-AG1 and FCN-AG2 model independently.

Automatic detection of COVID-19 in chest radiographs using serially concatenated deep and handcrafted features

Since the infectious disease occurrence rate in the human community is gradually rising due to varied reasons, appropriate diagnosis and treatments are essential to control its spread. The recently discovered COVID-19 is one of the contagious diseases, which infected numerous people globally. This contagious disease is arrested by several diagnoses and handling actions. Medical image-supported diagnosis of COVID-19 infection is an approved clinical practice. This research aims to develop a new Deep Learning Method (DLM) to detect the COVID-19 infection using the chest X-ray. The proposed work implemented two methods namely, detection of COVID-19 infection using (i) a Firefly Algorithm (FA) optimized deep-features and (ii) the combined deep and machine features optimized with FA. In this work, a 5-fold cross-validation method is engaged to train and test detection methods. The performance of this system is analyzed individually resulting in the confirmation that the deep feature-based technique helps to achieve a detection accuracy of > 92% with SVM-RBF classifier and combining deep and machine features achieves > 96% accuracy with Fine KNN classifier. In the future, this technique may have potential to play a vital role in testing and validating the X-ray images collected from patients suffering from the infection diseases.

Evaluation of an artificial intelligence (AI) system to detect tuberculosis on chest X-ray at a pilot active screening project in Guangdong, China in 2019

BACKGROUND: Although computer-aided detection (CAD) software employed with Artificial Intelligence (AI) system has been developed aiming to assist Tuberculosis (TB) triage, screening, and diagnosis, its clinical performance for tuberculosis screening remains unknown. OBJECTIVE: To evaluate performance of an CAD software for detecting TB on chest X-ray images at a pilot active TB screening project. METHODS: A CAD software scheme employed with AI was used to screen chest X ray images of participants and produce probability scores of cases being positive for TB. CAD-generated TB detection scores were compared with on-site and senior radiologists via several performance evaluation indices including area under the ROC curves (AUC), specificity, sensitive, and positive predict value. Pycharm CE and SPSS statistics software packages were used for data analysis. RESULTS: Among 2,543 participants, eight TB patients were identified from this screening pilot program. The AI-based CAD system outperformed the onsite (AUC = 0.740) and senior radiologists (AUC = 0.805) either using thresholds of 30% (AUC = 0.978) and 50% (AUC = 0.859) when taking the final diagnosis as the ground truth. CONCLUSIONS: The AI-based CAD software successfully detects all TB patients as identified from this study at a threshold of 30% . It demonstrates feasibility and easy accessibility to carry out large scale TB screening using this CAD software equipped in medical vans with chest X-ray imaging machine.

A new weighting scheme for arc based circle cone-beam CT reconstruction

In this paper, we present an arc based fan-beam computed tomography (CT) reconstruction algorithm by applying Katsevich’s helical CT image reconstruction formula to 2D fan-beam CT scanning data. Specifically, we propose a new weighting function to deal with the redundant data. Our weighting function ϖ ( x _ , λ ) is an average of two characteristic functions, where each characteristic function indicates whether the projection data of the scanning angle contributes to the intensity of the pixel x _ . In fact, for every pixel x _ , our method uses the projection data of two scanning angle intervals to reconstruct its intensity, where one interval contains the starting angle and another contains the end angle. Each interval corresponds to a characteristic function. By extending the fan-beam algorithm to the circle cone-beam geometry, we also obtain a new circle cone-beam CT reconstruction algorithm. To verify the effectiveness of our method, the simulated experiments are performed for 2D fan-beam geometry with straight line detectors and 3D circle cone-beam geometry with flat-plan detectors, where the simulated sinograms are generated by the open-source software “ASTRA toolbox.” We compare our method with the other existing algorithms. Our experimental results show that our new method yields the lowest root-mean-square-error (RMSE) and the highest structural-similarity (SSIM) for both reconstructed 2D and 3D fan-beam CT images.