Exploring the Low-Dose Limit for Focal Hepatic Lesion Detection with a Deep Learning-Based CT Reconstruction Algorithm: A Simulation Study on Patient Images.

This study aims to investigate the maximum achievable dose reduction for applying a new deep learning-based reconstruction algorithm, namely the artificial intelligence iterative reconstruction (AIIR), in computed tomography (CT) for hepatic lesion detection. A total of 40 patients with 98 clinically confirmed hepatic lesions were retrospectively included. The mean volume CT dose index was 13.66 ± 1.73 mGy in routine-dose portal venous CT examinations, where the images were originally obtained with hybrid iterative reconstruction (HIR). Low-dose simulations were performed in projection domain for 40%-, 20%-, and 10%-dose levels, followed by reconstruction using both HIR and AIIR. Two radiologists were asked to detect hepatic lesion on each set of low-dose image in separate sessions. Qualitative metrics including lesion conspicuity, diagnostic confidence, and overall image quality were evaluated using a 5-point scale. The contrast-to-noise ratio (CNR) for lesion was also calculated for quantitative assessment. The lesion CNR on AIIR at reduced doses were significantly higher than that on routine-dose HIR (all p < 0.05). Lower qualitative image quality was observed as the radiation dose reduced, while there were no significant differences between 40%-dose AIIR and routine-dose HIR images. The lesion detection rate was 100%, 98% (96/98), and 73.5% (72/98) on 40%-, 20%-, and 10%-dose AIIR, respectively, whereas it was 98% (96/98), 73.5% (72/98), and 40% (39/98) on the corresponding low-dose HIR, respectively. AIIR outperformed HIR in simulated low-dose CT examinations of the liver. The use of AIIR allows up to 60% dose reduction for lesion detection while maintaining comparable image quality to routine-dose HIR.

Deep Learning-Based Motion Correction in Projection Domain for Coronary Computed Tomography Angiography: A Clinical Evaluation.

OBJECTIVE: This study aimed to evaluate the clinical performance of a deep learning-based motion correction algorithm (MCA) in projection domain for coronary computed tomography angiography (CCTA). METHODS: A total of 192 patients who underwent CCTA examinations were included and divided into 2 groups based on the average heart rate (HR): group 1, 82 patients with HR of <75 beats per minute; group 2, 110 patients with HR of ≥75 beats per minute. The CCTA images were reconstructed with and without MCA. The subjective image quality was graded in terms of vessel visualization, sharpness, diagnostic confidence, and overall image quality using a 5-point scale, where cases with all scores of ≥3 were deemed interpretable. Objective image quality was measured through signal-to-noise ratio and contrast-to-noise ratio in regions relative to the vessels. The image quality scores for 2 reconstructions and effective dose between 2 groups were compared. RESULTS: The mean effective dose was similar between 2 groups. Neither group showed significant difference on objective image quality for 2 reconstructions. Images reconstructed with and without MCA were both found interpretable for group 1, whereas the subjective image quality was significantly improved by the MCA for all 4 metrics in group 2, with the interpretability increased from 80.91% to 99.09%. Compared with group 1, group 2 showed similar interpretability and diagnostic confidence, despite inferior overall image quality. CONCLUSIONS: In CCTA examinations, the deep learning-based MCA is capable of improving the image quality and diagnostic confidence for patients with increased HR to a similar level as for those with low HR.

Development of a CT image analysis-based scoring system to differentiate gastric schwannomas from gastrointestinal stromal tumors.

Purpose: To develop a point-based scoring system (PSS) based on contrast-enhanced computed tomography (CT) qualitative and quantitative features to differentiate gastric schwannomas (GSs) from gastrointestinal stromal tumors (GISTs). Methods: This retrospective study included 51 consecutive GS patients and 147 GIST patients. Clinical and CT features of the tumors were collected and compared. Univariate and multivariate logistic regression analyses using the stepwise forward method were used to determine the risk factors for GSs and create a PSS. Area under the receiver operating characteristic curve (AUC) analysis was performed to evaluate the diagnostic efficiency of PSS. Results: The CT attenuation value of tumors in venous phase images, tumor-to-spleen ratio in venous phase images, tumor location, growth pattern, and tumor surface ulceration were identified as predictors for GSs and were assigned scores based on the PSS. Within the PSS, GS prediction probability ranged from 0.60% to 100% and increased as the total risk scores increased. The AUC of PSS in differentiating GSs from GISTs was 0.915 (95% CI: 0.874-0.957) with a total cutoff score of 3.0, accuracy of 0.848, sensitivity of 0.843, and specificity of 0.850. Conclusions: The PSS of both qualitative and quantitative CT features can provide an easy tool for radiologists to successfully differentiate GS from GIST prior to surgery.

Deep learning-based motion correction algorithm for coronary CT angiography: Lowering the phase requirement for morphological and functional evaluation.

PURPOSE: To investigate the performance of a deep learning-based motion correction algorithm (MCA) at various cardiac phases of coronary computed tomography angiography (CCTA), and determine the extent to which it may allow for reliable morphological and functional evaluation. MATERIALS AND METHODS: The acquired image data of 53 CCTA cases, where the patient heart rate (HR) was ≥75 bpm, were reconstructed at 0, ±2, ±4, ±6, and ±8% deviations from each optimal systolic phase, with and without the MCA, yielding a total of 954 images (53 cases × 9 phases × 2 reconstructions). The overall image quality and diagnostic confidence were graded by two radiologists using a 5-point scale, with scores ≥3 being deemed clinically interpretable. Signal-to-noise ratio, contrast-to-noise ratio, vessel sharpness, and circularity were measured. The CCTA-derived fractional flow reserve (CT-FFR) was calculated in 38 vessels on 24 patients to identify functionally significant stenosis, using the invasive fractional flow reserve (FFR) as reference. All metrics were compared between two reconstructions at various phases. RESULTS: Inferior image quality was observed as the phase deviation was enlarged. However, MCA significantly improved the image quality at nonoptimal phases and the optimal phase. Coronary artery evaluation was feasible within 4% phase deviation using MCA, with interpretable overall image quality and high diagnostic confidence. With MCA, the performance of identifying functionally significant stenosis via CT-FFR was increased for images at various phase deviations. However, obvious decrease in accuracy, as compared to the image at the optimal phase, was found on those with deviations >4%. CONCLUSION: The deep learning-based MCA allows up to 4% phase deviation in acquiring CCTA for reliable morphological and functional evaluation on patients with high HRs.

Cross-modality image translation: CT image synthesis of MR brain images using multi generative network with perceptual supervision.

BACKGROUND: Computed tomography (CT) and magnetic resonance imaging (MRI) are the mainstream imaging technologies for clinical practice. CT imaging can reveal high-quality anatomical and physiopathological structures, especially bone tissue, for clinical diagnosis. MRI provides high resolution in soft tissue and is sensitive to lesions. CT combined with MRI diagnosis has become a regular image-guided radiation treatment plan. METHODS: In this paper, to reduce the dose of radiation exposure in CT examinations and ameliorate the limitations of traditional virtual imaging technologies, we propose a Generative MRI-to-CT transformation method with structural perceptual supervision. Even though structural reconstruction is structurally misaligned in the MRI-CT dataset registration, our proposed method can better align structural information of synthetic CT (sCT) images to input MRI images while simulating the modality of CT in the MRI-to-CT cross-modality transformation. RESULTS: We retrieved a total of 3416 brain MRI-CT paired images as the train/test dataset, including 1366 train images of 10 patients and 2050 test images of 15 patients. Several methods (the baseline methods and the proposed method) were evaluated by the HU difference map, HU distribution, and various similarity metrics, including the mean absolute error (MAE), structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), and normalized cross-correlation (NCC). In our quantitative experimental results, the proposed method achieves the lowest MAE mean of 0.147, highest PSNR mean of 19.27, and NCC mean of 0.431 in the overall CT test dataset. CONCLUSIONS: In conclusion, both qualitative and quantitative results of synthetic CT validate that the proposed method can preserve higher similarity of structural information of the bone tissue of target CT than the baseline methods. Furthermore, the proposed method provides better HU intensity reconstruction for simulating the distribution of the CT modality. The experimental estimation indicates that the proposed method is worth further investigation.

[Clinical Application of "Three-Low" Technique Combined with Artificial Intelligence Iterative Reconstruction Algorithm in Aortic CT Angiography].

Objective: To explore the application value of the "three-low" technique (low radiation dose, low contrast agent dosage and low contrast agent flow rate) combined with artificial intelligence iterative reconstruction (AIIR) in aortic CT angiography (CTA). Methods: A total of 33 patients who underwent aortic CTA were prospectively enrolled. Based on the time of their follow-up examinations, the imaging data were divided into Group A and Group B, with Group A being the control group (100 kV, 0.8 mL/kg, 5 mL/s) and Group B being the "three-low" technique group (70 kV, 0.5 mL/kg, 3 mL/s). In group A, the images were reconstructed by Karl iterative algorithm. Group B was divided into B1 and B2 subgroups, with their images being reconstructed by Karl iterative algorithm and AIIR, respectively. The CT and SD values of the ascending aorta, descending aorta, abdominal aorta, left common iliac artery and right common iliac artery were measured, and the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated. The subjective scoring of image quality was performed. The radiation dose parameters were documented. Results: Differences in the CT value, SD value, SNR and CNR of the three groups were statistically significant ( P<0.001). The CT value, SNR and CNR of group B2 were significantly higher than those of group B1, while the SD value of group B2 was significantly lower than that of group B1 ( P<0.017). There was no significant difference between the CT values of group A and those of group B2 ( P>0.017). The SD values, SNR and CNR in group B2 were better than those in group A ( P>0.017). There was significant difference in the subjective evaluation of image quality among the three groups ( P<0.05), but there was no significant difference between group A and group B2 ( P>0.017). The radiation dose and contrast medium dosage in group B decreased 84.14% and 37.08%, respectively, compared with those of group A. Conclusion: With the "three-low" technique combined with AIIR algorithm, the image quality of aortic CTA obtained is comparable to that of conventional dose scanning, while the radiation dose, contrast agent dosage and contrast agent flow rate of patients are significantly reduced.

Image quality assessment of artificial intelligence iterative reconstruction for low dose aortic CTA: A feasibility study of 70 kVp and reduced contrast medium volume.

PURPOSE: To investigate the image quality and feasibility of a novel artificial intelligence iterative reconstruction (AIIR) algorithm for aortic computer tomography angiography (CTA) with a low radiation dose and contrast material (CM) dosage protocol in comparison with hybrid iterative reconstruction (HIR) algorithm for standard-of-care aortic CTA. METHODS: Fifty consecutive patients (mean age 58 ± 14 years, mean BMI 24.5 ± 4.7 kg/m2) with aortic diseases were prospectively enrolled. All patients underwent at least twice follow-up aortic CTA examinations. Standard dose CT (SDCT) was applied in the initial follow-up examination (100 kVp, auto mAs, contrast dose 0.8 mgL/kg), images were reconstructed with HIR (SDCT-HIR). In the second follow-up examination, patients underwent scanning with low dose CT (LDCT) (70 kVp, auto mAs, contrast dose 0.5 mgL/kg), images were reconstructed with HIR (LDCT-HIR) as well as AIIR (LDCT-AIIR). Attenuation values, noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were measured for objective analysis. Subjective image quality was rated by two blinded radiologists using a 5-point scale. The effective radiation dose and CM dosage were also recorded. RESULTS: The effective radiation dose (1.58 ± 0.17 mSv vs. 9.96 ± 1.05 mSv, P < 0.001) and CM dosage (34.38 ± 5.43 ml vs. 54.64 ± 8.63 ml, P < 0.001) achieved a remarkable reduction of 84.14% and 37.08% in the LDCT compared to the SDCT. The attenuation was similar among the three reconstructed images (P > 0.05). Compared to LDCT-HIR images, LDCT-AIIR showed a lower noise and higher SNR and CNR. For qualitative analysis, there were no significant differences between the LDCT-AIIR and the SDCT-HIR images among four metrics (P > 0.05). CONCLUSIONS: Compared to standard-of-care aortic CTA with HIR, the application of the AIIR algorithm allows for radiation dose and CM dosage reduction while preserving image quality on low dose aortic CTA.

Inverted Pyramid Morphology Control by Acid Modification and Application for PERC Solar Cells.

Silicon inverted pyramid (IP) structures, with lower reflectance and increased surface recombination, are one of the best choices for light-trapping structures of high-efficiency silicon solar cells. The solution process of IP generally goes through three main steps: porous silicon etched by metal-assisted chemical etching, acid etching, and alkali anisotropic etching. In this paper, the role that acid modification plays in IP preparation and the application of our optimized texture for passivated emitter and rear solar cells (PERC) were investigated. Experimental results show that acid plays a decisive role in optimizing and modifying the morphology of porous silicon; thus, the morphology of porous silicon has no direct influence on the morphology of IP. In addition, the opening size of IP is mainly determined by the size of silicon micron holes modified by the acid process. PC1D simulation results manifest that IPs can increase the short-circuit current density (J sc) of devices by 1.04 mA/cm2 and power conversion efficiency by 0.55%; hence, our optimized IP-based PERC achieve the highest simulative conversion efficiency of 23.21%. This is an effective and important way to manipulate the structure of IP, which points out the direction of fabrication and application of high-efficiency IP textures.

Diagnostic performance of corrected FFRCT metrics to predict hemodynamically significant coronary artery stenosis.

OBJECTIVES: To determine the diagnostic performance of the fractional flow reserve (FFR) derived from coronary computed tomography angiography (CCTA) (FFRCT) difference across the lesion (ΔFFRCT lesion) or the vessel (ΔFFRCT vessel) and the gradient of FFRCT for the identification of hemodynamically significant coronary stenosis. METHODS: From June 2016 to December 2018, 73 patients suspected of having coronary artery disease who underwent CCTA followed invasive coronary angiography (ICA) within 1 month were retrospectively included. ΔFFRCT lesion, ΔFFRCT vessel, and FFRCT gradient were calculated. Performance characteristics of different corrected FFRCT metrics in detecting ischemic stenosis were analyzed. Impacts of coronary calcification and lesion length on the corrected FFRCT metrics were also analyzed. RESULTS: The diagnostic sensitivities, specificities, and accuracies of 94.4%, 88.7%, and 91.0% with ΔFFRCT lesion, 57.1%, 72.3%, and 65.2% with ΔFFRCT vessel, and 50.0%, 85.1%, and 68.5% with FFRCT gradient, respectively, were detected. There was higher specificity, accuracy, and area under the curve (AUC) for ΔFFRCT lesion compared with CCTA (p < 0.05 for all). The specificity and AUC of FFRCT gradient and ΔFFRCT vessel were significantly higher than CCTA (p < 0.05 for all). Coronary calcification showed no impact on corrected FFRCT metrics. ΔFFRCT lesion for lesion length ratio (LLR) < 1/10 was significantly lower than that for LLR 1/10 to 3/10 and LLR > 3/10. CONCLUSIONS: ΔFFRCT lesion was significantly correlated with the hemodynamically significant coronary artery stenosis. ΔFFRCT lesion had the potential to be immediately used in real-world practice to discriminate ischemic coronary artery stenosis. KEY POINTS: • The difference of FFRCT across the lesion or the vessel and the gradient of FFRCT was related to the hemodynamically significant coronary artery stenosis. • The difference of FFRCT across the lesion showed the best diagnostic performance in detecting the hemodynamically significant coronary artery stenosis. • Coronary calcification showed no impact on corrected FFRCT metrics, while lesion length related to the difference of FFRCT across the lesion.

Realization of Quasi-Omnidirectional Solar Cells with Superior Electrical Performance by All-Solution-Processed Si Nanopyramids.

Large-scale (156 mm × 156 mm) quasi-omnidirectional solar cells are successfully realized and featured by keeping high cell performance over broad incident angles (θ), via employing Si nanopyramids (SiNPs) as surface texture. SiNPs are produced by the proposed metal-assisted alkaline etching method, which is an all-solution-processed method and highly simple together with cost-effective. Interestingly, compared to the conventional Si micropyramids (SiMPs)-textured solar cells, the SiNPs-textured solar cells possess lower carrier recombination and thus superior electrical performances, showing notable distinctions from other Si nanostructures-textured solar cells. Furthermore, SiNPs-textured solar cells have very little drop of quantum efficiency with increasing θ, demonstrating the quasi-omnidirectional characteristic. As an overall result, both the SiNPs-textured homojunction and heterojunction solar cells possess higher daily electric energy production with a maximum relative enhancement approaching 2.5%, when compared to their SiMPs-textured counterparts. The quasi-omnidirectional solar cell opens a new opportunity for photovoltaics to produce more electric energy with a low cost.

Superior broadband antireflection from buried Mie resonator arrays for high-efficiency photovoltaics.

Establishing reliable and efficient antireflection structures is of crucial importance for realizing high-performance optoelectronic devices such as solar cells. In this study, we provide a design guideline for buried Mie resonator arrays, which is composed of silicon nanostructures atop a silicon substrate and buried by a dielectric film, to attain a superior antireflection effect over a broadband spectral range by gaining entirely new discoveries of their antireflection behaviors. We find that the buried Mie resonator arrays mainly play a role as a transparent antireflection structure and their antireflection effect is insensitive to the nanostructure height when higher than 150 nm, which are of prominent significance for photovoltaic applications in the reduction of photoexcited carrier recombination. We further optimally combine the buried Mie resonator arrays with micron-scale textures to maximize the utilization of photons, and thus have successfully achieved an independently certified efficiency of 18.47% for the nanostructured silicon solar cells on a large-size wafer (156 mm × 156 mm).

A simple strategy for synthesizing highly luminescent carbon nanodots and application as effective down-shifting layers.

We propose a novel strategy to prepare highly luminescent carbon nanodots (C-dots) by employing a hydrothermal method with citric acid as the carbon source and ethylenediamine as the nitrogen source, together with adding moderate ammonia water (AW) to achieve both appropriate inner structure and excellent N passivation. The effect of pH value and AW amount on the luminescence properties has been thoroughly investigated. The photoluminescence quantum yield of the resultant C-dots reaches as high as 84.8%, which is of 10.56% higher than that of the C-dots synthesized in the absence of AW in the reaction precursors. We have further combined the highest luminescent C-dots with polyvinyl alcohol to form luminescent down-shifting layers on silicon nanowire solar cells. An effective enhancement of short-circuit current density has been realized and the contribution of the down-shifting has been extracted quantitatively from the deterioration of surface reflectance and the gain of the optical absorption redistribution by means of a theoretical model on external quantum efficiency analysis.

High-efficiency nanostructured silicon solar cells on a large scale realized through the suppression of recombination channels.

Nanostructured silicon solar cells show great potential for new-generation photovoltaics due to their ability to approach ideal light-trapping. However, the nanofeatured morphology that brings about the optical benefits also introduces new recombination channels, and severe deterioration in the electrical performance even outweighs the gain in optics in most attempts. This Research News article aims to review the recent progress in the suppression of carrier recombination in silicon nanostructures, with the emphasis on the optimization of surface morphology and controllable nanostructure height and emitter doping concentration, as well as application of dielectric passivation coatings, providing design rules to realize high-efficiency nanostructured silicon solar cells on a large scale.