Image Quality Measures Research Articles

Multi-vendor validation of automated left ventricular analysis from 3D echocardiography using artificial intelligence

Abstract Background Accurate echocardiographic assessment of left ventricular (LV) function is dependent on both image acquisition and analysis. Although 3D echocardiography (3DE) is increasingly recommended for volume quantification, manually derived volumes are limited by operator subjectivity, variable image quality, and the lengthy duration of analysis. Consequently, several automated approaches based on artificial intelligence (AI) have been proposed to segment the LV cavity from 3DE and have demonstrated better accuracy and reproducibility compared to human experts. While the utility of AI models in the clinic depends on their ability to generalise to different populations and ultrasound vendors, the external validation of developed models remains limited. Purpose This study evaluates the ability of an existing AI model trained using data from a single vendor to accurately estimate routine 3DE LV indices in an independent external dataset acquired using equipment from a different vendor. Methods A previously published and internally validated deep learning model for segmentation of the LV cavity and myocardium from 3DE was used to automatically derive LV indices including end-diastolic volume (EDV), end-systolic volume (ESV), LV mass, and ejection fraction (EF) [1]. A second independent dataset consisting of paired 3DE and CMR images from 65 marathon runners from a different centre (using equipment from a different vendor) was subsequently used for external validation. Images from both datasets were subjectively scored from 1 (poor) to 5 (excellent) as a measure of image quality. Agreement with CMR as the reference was assessed using a two-way mixed effects intraclass correlation coefficient (ICC), supplemented by Bland-Altman analysis, to determine the model bias (between 3DE and CMR) and 95% limits of agreement (LOA). To assess whether performance could be improved by inclusion of multi-vendor data during training, the initial model was fine-tuned using a small subset (n=9) of images from the second centre. Results The mean image quality score from the second dataset was 2.5 (± 1.1), which was significantly lower than the mean of 3.4 (± 1.0) from the first dataset. Despite vendor and image quality differences, the initial model produced good agreement with CMR in EDV, ESV, and LV mass but only poor agreement in EF (Fig 1). Fine tuning greatly improved agreement in EF with reduced bias and narrower 95% LOA (Fig 2). Conclusion With the addition of a small number of cases for fine-tuning, AI models for automated LV analysis can achieve good performance analysing 3DE images acquired from a different vendor and population. Advancing model generalisability through more sophisticated data augmentation strategies may enable truly vendor-agnostic automated LV analysis from 3DE.

Pediatric contrast-enhanced chest CT on a photon-counting detector CT: radiation dose and image quality compared to energy-integrated detector CT.

Photon counting detector (PCD) CT benefits from reduced noise compared with conventional energy-integrating detector (EID) CT, which should translate to improved image quality and reduced radiation exposure for pediatric patients undergoing chest CT with IV contrast. To determine the differences in radiation exposure and image quality of PCD CT and EID CT in pediatric chest CT with intravenous (IV) contrast. In this institutional review board-approved retrospective observational study, 20 scan pairs (20 PCD CT; 20 EID CT) for children who underwent chest CT with IV contrast on both a PCD CT (Siemens NAEOTOM Alpha) and an EID CT (Siemens SOMATOM Definition Edge or Force) within 12months were reviewed independently by three pediatric radiologists for three subjective quality features on 5-point Likert scales: overall quality, small structure delineation, and motion artifact. Objective measures of image quality (image noise, signal-to-noise ratio, and contrast-to-noise ratio) were assessed by a single radiologist in several locations in the chest through region of interest measurement of Hounsfield units (HU) and standard deviation. Patient-related and radiation exposure parameters were collected for each scan and summarized with median and interquartile range (IQR). The Wilcoxon rank-sum test was utilized to compare groups. A P < 0.05 indicated statistical significance. Inter-observer agreement of subjective image quality metrics was analyzed using weighted kappa. Age (14.2years vs 13.8years, P= 0.15), height (P= 0.13), weight (P= 0.21), and BMI (P = 0.24) did not significantly differ between groups. There were 10 male and 3 female patients. Compared to EID CT, PCD CT showed lower radiation exposure parameters including volumetric CT dose index, 1.7mGy (IQR 1.1-2.4mGy) vs 3.8mGy (IQR 2.0-4.7mGy) (P< 0.01), and size-specific dose estimate, 2.6mGy (IQR 1.8-3.1mGy) vs 5.0mGy (IQR 3.3-6.2mGy) (P< 0.01). Objective image quality of lung parenchyma was improved on the PCD CT scanner, including image noise 119.5 HU (IQR 95.4-135.7 HU) vs 143.1 HU (IQR 125.4-169.8 HU) (P < 0.01), signal-to-noise ratio (SNR) -6.1 (IQR -8.4 to -4.8) vs -4.9 (IQR -5.6 to -3.8) (P= 0.01), and contrast-to-noise ratio -63.9 (-84.1 to -57.5) vs -60.5 (-76.3 to -52.5) (P = 0.01). Motion artifact was improved on the PCD CT scanner (P< 0.01). No significant differences in overall image quality or small structure delineation were identified (P= 0.06 and P= 0.31). PCD CT pediatric chest CT had significantly reduced radiation exposure, improved image quality, and reduced motion artifact compared with EID CT.

High on sparsity: Interbin compensation of cardiac motion for improved assessment of left-ventricular function using 5D whole-heart MRI.

Cardiac magnetic resonance is the gold standard for evaluating left-ventricular ejection fraction (LVEF). Standard protocols, however, can be inefficient, facing challenges due to significant operator and patient involvement. Although the free-running framework (FRF) addresses these challenges, the potential of the extensive data it collects remains underutilized. Therefore, we propose to leverage the large amount of data collected by incorporating interbin cardiac motion compensation into FRF (FRF-MC) to improve both image quality and LVEF measurement accuracy, while reducing the sensitivity to user-defined regularization parameters. FRF-MC consists of several steps: data acquisition, self-gating signal extraction, deformation field estimations, and motion-resolved reconstruction with interbin cardiac motion compensation. FRF-MC was compared with the original 5D-FRF method using LVEF and several image-quality metrics. The cardiac regularization weight ( ) was optimized for both methods by maximizing image quality without compromising LVEF measurement accuracy. Evaluations were performed in numerical simulations and in 9 healthy participants. In vivo images were assessed by blinded expert reviewers and compared with reference standard 2D-cine images. Both in silico and in vivo results revealed that FRF-MC outperformed FRF in terms of image quality and LVEF accuracy. FRF-MC reduced temporal blurring, preserving detailed anatomy even at higher cardiac regularization weights, and led to more accurate LVEF measurements. Optimized produced accurate LVEF for both methods compared with the 2D-cine reference (FRF-MC: 0.59% [-7.2%, 6.0%], p = 0.47; FRF: 0.86% [-8.5%, 6.7%], p = 0.36), but FRF-MC resulted in superior image quality (FRF-MC: 2.89 ± 0.58, FRF: 2.11 ± 0.47; p < 10-3). Incorporating interbin cardiac motion compensation significantly improved image quality, supported higher cardiac regularization weights without compromising LVEF measurement accuracy, and reduced sensitivity to user-defined regularization parameters.

Optimized underwater light attenuation prior-based depth estimation and adaptive feature fusion CNN for underwater image and video enhancement

Underwater images and videos play an essential communication tool in exploring ocean resources and understanding underwater scene perception. The underwater imaging environment and the complex lighting conditions degrade the quality of the images and make them low contrast, color deviation, and blurring. These degraded images lack effective target recognition information, significantly impacting the performance of underwater application systems. To solve the above issues, a new module for underwater video as well as image enhancement is proposed. The input video is given to the frame extraction phase for converting the video into the frame. Thereafter, the blur detection and determination are done using Laplacian's variance technique for determining blurred images. The depth estimation is done using Underwater Light Attenuation Prior (ULAP), wherein coefficients are determined using the proposed Competitive Multi-Verse bird swarm Optimization (CMVBSO). The CMVBSO is combined by integrating a competitive multiverse optimizer (CMVO) and a Bird Swarm Algorithm (BSA). The blurred value, input video frame and depth estimated output are given to adaptive feature fusion CNN in which training is done utilizing Manta-Ray Foraging Lion Optimization (MRLLO). The pixel enhancement is carried out using a Type II Fuzzy system and Cuckoo Search optimization algorithm filter (T2FCS) filter. The proposed CMVBSO_ULAP provided better performance with a Peak signal to noise ratio (PSNR) of 35.037 dB, Mean square error (MSE) of 3.684, Structural similarity index (SSIM) of 0.911, underwater image quality measure (UIQM) of 5.266 and Underwater Color Image Quality Evaluation (UCIQE) of 0.818 respectively.

Design of Prototype Phantom to Measure Quality Control for Conventional X-ray Machine Using Quantitative Image Parameters

A variety of phantoms that measure image quality parameters are commercially expensive and often complicated to use as they measure only one parameter of a particular x-ray machine, which leads to the difficulty of measuring the quality of x-ray image and thus increase the risk of radiation to the patient and society. The main objective of this study was to design a prototype phantom to measure quality control for conventional x-ray machine using quantitative image parameters. In this study, 5 conventional x-ray machines were investigated using wire phantom designed by the researcher, which consisted of variable thicknesses used to test variable exposure factors. This study introduced a more reliable method of measuring the resolution, which is modulation transfer function, which gives a complete description of the resolution instead of using full width at half maximum or the visibility method, which is more qualitative where modulation transfer function is a real quantitative method, by designing a prototype phantom that consisted of 5 wires with different thicknesses and kVp for 5 x-ray units and assessing the diagnostic x-ray machine resolution by the optimum kV found relative to machine type. This study showed that the one with the best machine resolution was Shimadzu (2011) for Khartoum Emergency Hospital, which had a high resolution of 97% at 46 kV for thickness of 1.4 mm compared with Toshiba 2011 for Almotkamil Hospital, which had 92% at 46 kVp. Also, the result showed that for the 2 types of x-ray machines the x-ray tubes do not produce the same exposure, and the output decreased with age of the x-ray unit; also, the resolution reduced when the thickness of the wires decreased.

Clinical Validation Study of Deep Learning-Generated Magnetic Resonance Images

This research utilizes a deep learning-based image generation algorithm to generate pseudo-sagittal STIR sequences from sagittal T1WI and T2WI MR images. The evaluations include both subjective assessments by two physicians and objective analyses, measuring image quality through SNR and CNR in ROIs of five different tissues. Further analyses, including MAE, PSNR, SSIM, and COR, establish a strong correlation between the generated STIR sequences and the gold standard, with Bland-Altman analysis indicating pixel consistency. The findings indicate that the deep learning-generated STIR sequences not only align with but potentially surpass the gold standard in terms of image quality and clinical diagnostic capabilities. Moreover, the approach demonstrates promise for clinical implementation, offering reduced scan time and enhanced imaging efficiency.

Investigating the use of signal detection information in supervised learning-based image denoising with consideration of task-shift.

Recently, learning-based denoising methods that incorporate task-relevant information into the training procedure have been developed to enhance the utility of the denoised images. However, this line of research is relatively new and underdeveloped, and some fundamental issues remain unexplored. Our purpose is to yield insights into general issues related to these task-informed methods. This includes understanding the impact of denoising on objective measures of image quality (IQ) when the specified task at inference time is different from that employed for model training, a phenomenon we refer to as "task-shift." A virtual imaging test bed comprising a stylized computational model of a chest X-ray computed tomography imaging system was employed to enable a controlled and tractable study design. A canonical, fully supervised, convolutional neural network-based denoising method was purposely adopted to understand the underlying issues that may be relevant to a variety of applications and more advanced denoising or image reconstruction methods. Signal detection and signal detection-localization tasks under signal-known-statistically with background-known-statistically conditions were considered, and several distinct types of numerical observers were employed to compute estimates of the task performance. Studies were designed to reveal how a task-informed transfer-learning approach can influence the tradeoff between conventional and task-based measures of image quality within the context of the considered tasks. In addition, the impact of task-shift on these image quality measures was assessed. The results indicated that certain tradeoffs can be achieved such that the resulting AUC value was significantly improved and the degradation of physical IQ measures was statistically insignificant. It was also observed that introducing task-shift degrades the task performance as expected. The degradation was significant when a relatively simple task was considered for network training and observer performance on a more complex one was assessed at inference time. The presented results indicate that the task-informed training method can improve the observer performance while providing control over the tradeoff between traditional and task-based measures of image quality. The behavior of a task-informed model fine-tuning procedure was demonstrated, and the impact of task-shift on task-based image quality measures was investigated.

Simultaneous multislice echo-planar diffusion-weighted imaging (DWI) in patients with focal liver lesions: a comparative study with conventional DWI.

Simultaneous multislice (SMS) technology improves acquisition efficiency of diffusion-weighted imaging (DWI). This study aimed to evaluate the performance of SMS-DWI in image quality and apparent diffusion coefficient (ADC) measurements for focal liver lesions (FLLs) as compared with that of conventional DWI (CON-DWI). The institutional ethics committee of West China Hospital, Sichuan University approved this single-center, prospective study conducted from February 2021 to March 2022. Free-breathing SMS-DWI and CON-DWI examinations were acquired on a 3-T scanner with b-values of 50, 400, and 800 s/mm2. Qualitative image quality and quantitative measurements of signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and ADC were compared between SMS-DWI and CON-DWI. The ADC values for FLLs were further compared between SMS-DWI and CON-DWI in different patient subgroups. The intra- and interreader agreements were assessed. Significance was set at P<0.05. This study included 116 patients (96 males, 20 females; mean age 52.0±10.7 years) with 119 FLLs. No significant differences were observed between SMS-DWI and CON-DWI regarding overall image quality in any b-value DWIs, and there were also no differences observed between SMS-DWI and CON-DWI (b=800 s/mm2) for either SNR or CNR (both P values >0.05). ADC values obtained from CON-DWI were higher than those from SMS-DWI in all FLLs [(1.31±0.47)×10-3 vs. (1.26±0.46)×10-3 mm2/s; P=0.004], and similar findings were observed across the different patient subgroups. The consistency analysis showed intrareader intraclass correlation coefficient (ICC) values of 0.792-0.944 and interreader ICC values of 0.758-0.861 for quantitative measurements (SNR, CNR, and ADC) and kappa values of 0.609-0.878 for qualitative image quality. SMS-DWI achieved a 37% reduction in scan time compared to CON-DWI while maintaining comparable overall image quality. Notably, the ADC values for FLLs were observed to be quantitatively lower with SMS-DWI.

Noncontrast free-breathing ECG-gated 3D balanced steady-state free precession in congenital heart disease and aortopathy evaluation.

High-fidelity cardiac magnetic resonance (MR) imaging plays a pivotal role in the surveillance of congenital heart disease (CHD) and aortopathy. We aimed to evaluate the quality and accuracy of free breathing, ECG-gated noncontrast three-dimensional (3D) balanced steady-state free precession (bSSFP) in cases of CHDs and aortopathies using contrast-enhanced 3D bSSFP as a reference. We also used one of our routinely used non-ECG-gated 2D-single-shot (SSh) bSSFP sequence as an adjunct to noncontrast 3D bSSFP. Institutional review board approval was obtained to perform a systematic retrospective analysis of image quality and vascular measurements. Patients with CHD and aortopathy, who were undergoing clinically indicated contrast-enhanced 3D bSSFP, were prospectively identified to also undergo additional noncontrast 3D bSSFP and 2D SSh bSSFP imaging as part of a clinical quality improvement initiative aimed at reducing the use of contrast when feasible. Two readers, blinded to each other's evaluations, graded image quality on a 5-point Likert scale and performed vascular measurements in separate sessions for both 3D bSSFP images. They also reported the visibility of various mediastinal great vessels on 2D SSh bSSFP images. Raw agreement, the weighted kappa statistic, and intra-class correlation coefficients (ICCs) were computed to assess the consistency and agreement between the two readers. Comparative analysis of noncontrast and contrast-enhanced 3D bSSFP imaging was performed in adult and pediatric patients using a two-sided paired t-test and Bland-Altman analysis. A P-value < 0.05 was considered significant for all inference testing. A total of 29 patients (17 males, median age 20.3years, interquartile range (IQR) 12.5, age range 7-39years), including 11 pediatric patients under the age of 18years (6 males, median age 14.5years, IQR 4.0, age range 7-17years), underwent retrospective analysis. The overall image quality score for contrast-enhanced 3D bSSFP was significantly higher (P < 0.0001) than that of noncontrast 3D bSSFP for both all subjects (4.4 ± 0.2, range 4.0-4.9 vs 3.7 ± 0.4, range 3.1-4.7) and only pediatric subjects (4.3 ± 0.3, range 4.0-4.9 vs 3.6 ± 0.5, range 3.1-4.4). By combining noncontrast 3D bSSFP and 2D bSSFP, reader 1 and reader 2 rated 423 and 420 vessels diagnostic, respectively, in a total of 435 vessel segments. All landmarks showed similar mean vessel diameters without significant differences between noncontrast and contrast-enhanced 3D bSSFP MR angiography (r = 0.99, bias -0.31mm, 95% limits of agreement -2.04mm to 1.43mm). Although contrast-enhanced images had better overall image quality, an imaging protocol consisting of noncontrast 2D SSh bSSFP and 3D bSSFP whole-chest images provides diagnostically adequate image quality, and accurate vascular measurements, comparable to free-breathing contrast-enhanced 3D bSSFP in both children and adults with CHD and aortopathies.

Underwater image enhancement using contrast correction

AbstractLight‐induced degeneration of underwater images occurs by physical features of seawater. According to the wavelength of the colour spectrum, light reduces intensity significantly when it moves through water. The greatest wavelength of light that is visible gets absorbed first. Red and blue absorb the most and least, respectively. Because of the reducing consequences of the light spectrum, underwater images having poor contrast can be obtained. As a result, the crucial data contained inside these images will not be effectively retrieved for later analysis. The recent research suggests a novel approach to enhance the contrast while decreasing noise in underwater images. The recommended approach involves image histogram transformation using two significant colour spaces, Red‐Green‐Blue (RGB) and Hue‐Saturation‐Value (HSV). The histogram of the dominant colour channel (blue channel) in the RGB colour model is extended towards the lower level, containing a maximum limitation of 95%, while the inferior red colour channel has been extended towards the upper side, containing a minimum limitation of 5%. During the entire dynamic range, the green colour channel having the dominant and inferior colour channels expands in each direction. The Rayleigh distribution has been utilized for developing various stretching actions within the RGB colour space. The image has been converted to the HSV colour space, having the S and V elements adjusted within 1% of their minimum and maximum values. The suggested approach is examined in both qualitative and quantitative analysis. According to qualitative analysis, the recommended approach substantially boosts image contrast, lowers its blue and green effect, and minimizes over‐enhanced and under‐enhanced sections in the final resultant underwater image. The quantitative examination of 500 large scale underwater images dataset reveals that the suggested technique generates better results. The dataset images are grouped into small fish images, blue coral images, stone wall images, and coral branch images. The quantitative examination of all these four groups have been evaluated and shown. The average mean square error, peak signal to noise ratio, underwater image quality measurement, and underwater colour image quality evaluation values of dataset images are 76.69, 31.25, 3.85, and 0.64, respectively. These values of our proposed work outperform six other previous methods.

Task-Adaptive Angle Selection for Computed Tomography-Based Defect Detection.

Sparse-angle X-ray Computed Tomography (CT) plays a vital role in industrial quality control but leads to an inherent trade-off between scan time and reconstruction quality. Adaptive angle selection strategies try to improve upon this based on the idea that the geometry of the object under investigation leads to an uneven distribution of the information content over the projection angles. Deep Reinforcement Learning (DRL) has emerged as an effective approach for adaptive angle selection in X-ray CT. While previous studies focused on optimizing generic image quality measures using a fixed number of angles, our work extends them by considering a specific downstream task, namely image-based defect detection, and introducing flexibility in the number of angles used. By leveraging prior knowledge about typical defect characteristics, our task-adaptive angle selection method, adaptable in terms of angle count, enables easy detection of defects in the reconstructed images.

Task-based assessment for neural networks: evaluating undersampled MRI reconstructions based on human observer signal detection.

Recent research explores using neural networks to reconstruct undersampled magnetic resonance imaging. Because of the complexity of the artifacts in the reconstructed images, there is a need to develop task-based approaches to image quality. We compared conventional global quantitative metrics to evaluate image quality in undersampled images generated by a neural network with human observer performance in a detection task. The purpose is to study which acceleration (2×, 3×, 4×, 5×) would be chosen with the conventional metrics and compare it to the acceleration chosen by human observer performance. We used common global metrics for evaluating image quality: the normalized root mean squared error (NRMSE) and structural similarity (SSIM). These metrics are compared with a measure of image quality that incorporates a subtle signal for a specific task to allow for image quality assessment that locally evaluates the effect of undersampling on a signal. We used a U-Net to reconstruct under-sampled images with 2×, 3×, 4×, and 5× one-dimensional undersampling rates. Cross-validation was performed for a 500- and a 4000-image training set with both SSIM and MSE losses. A two-alternative forced choice (2-AFC) observer study was carried out for detecting a subtle signal (small blurred disk) from images with the 4000-image training set. We found that for both loss functions, the human observer performance on the 2-AFC studies led to a choice of a 2× undersampling, but the SSIM and NRMSE led to a choice of a 3× undersampling. For this detection task using a subtle small signal at the edge of detectability, SSIM and NRMSE led to an overestimate of the achievable undersampling using a U-Net before a steep loss of image quality between 2×, 3×, 4×, 5× undersampling rates when compared to the performance of human observers in the detection task.

Enhancing quality measurement for visible and invisible watermarking based on M-SVD and DCT

This study introduces an advanced method for evaluating non-blind watermarking quality, leveraging both visible and invisible watermarking techniques grounded in principles of discrete cosine transform (DCT) and modified singular value decomposition (M-SVD). The primary focus is to refine the assessment process of watermarked images by integrating M-SVD, known for its efficacy in measuring image quality and watermarking performance. Results from the M-SVD implementation exhibit a striking resemblance to the original images. The mean squared error (MSE) values for watermarked images range from 0.0003 to 0.0168, while peak signal-to-noise ratio (PSNR) values vary between 42.52 dB and 82.72 dB. These outcomes underscore the potential of DCT and M-SVD techniques in bolstering watermarking processes, especially in invisible watermarking contexts.

64Cu]Copper chloride PET-CT: a comparative evaluation of fasting and non-fasting states in patients of prostate carcinoma.

Altered copper metabolism in cancer has been linked to increased intracellular copper uptake mediated by human copper transporter 1, with [64Cu]Cu2+ as a potential biomarker for cancer theranostics. [64Cu]CuCl2 PET-CT though explored in various malignancies, a lack of standardized protocol exists, particularly regarding fasting status before imaging. This analysis aimed to evaluate the requirement of fasting for [64Cu]CuCl2 PET-CT along with temporal changes in physiological organ uptake in delayed scans. A total of 26 patients of prostate carcinoma who underwent [64Cu]CuCl2 PET-CT imaging were divided into two groups: (1) nonfasting (n = 12) and (2) fasting (n = 14). The nonfasting group received an average dose of 350 MBq, while the fasting group received 300 MBq of [64Cu]CuCl2, and PET-CT images acquired approximately 60-90 min (1 h image) and 3-3.5 h (delayed image) after intravenous injection of the tracer. An experienced nuclear medicine physician evaluated the images for qualitative assessment between the groups. Multiple spherical regions of interest were placed at sites of physiological organ uptake of the tracer and over the diseased lesions to measure the mean SUVmax. No significant difference was observed in the qualitative assessment of the images between the two groups (except for a slight predilection towards more hepatic tracer retention observed in the fasting group), including in the delayed images. The liver demonstrated the highest tracer uptake in all patients, with a mean SUVmax of 21.5 in the fasting group and 19.7 in the nonfasting group, showing no significant difference (P = 0.32). The kidneys, intestines, and salivary glands also showed similar trends of tracer uptake in both groups. The study illustrated that the fasting or nonfasting status did not affect image quality or semiquantitative measurements significantly in physiological organs and diseased lesions in patients with carcinoma prostate.

Impact on Image Quality and Diagnostic Performance of Dual-Layer Detector Spectral CT for Pulmonary Subsolid Nodules: Comparison With Hybrid and Model-Based Iterative Reconstruction.

To evaluate the image quality and diagnostic performance of pulmonary subsolid nodules on conventional iterative algorithms, virtual monoenergetic images (VMIs), and electron density mapping (EDM) using a dual-layer detector spectral CT (DLSCT). This retrospective study recruited 270 patients who underwent DLSCT scan for lung nodule screening or follow-up. All CT examinations with subsolid nodules (pure ground-glass nodules [GGNs] or part-solid nodules) were reconstructed with hybrid and model-based iterative reconstruction, VMI at 40, 70, 100, and 130 keV levels, and EDM. The CT number, objective image noise, signal-to-noise ratio, contrast-to-noise ratio, diameter, and volume of subsolid nodules were measured for quantitative analysis. The overall image quality, image noise, visualization of nodules, artifact, and sharpness were subjectively rated by 2 thoracic radiologists on a 5-point scale (1 = unacceptable, 5 = excellent) in consensus. The objective image quality measurements, diameter, and volume were compared among the 7 groups with a repeated 1-way analysis of variance. The subjective scores were compared with Kruskal-Wallis test. A total of 198 subsolid nodules, including 179 pure GGNs, and 19 part-solid nodules were identified. Based on the objective analysis, EDM had the highest signal-to-noise ratio (164.71 ± 133.60; P < 0.001) and contrast-to-noise ratio (227.97 ± 161.96; P < 0.001) among all image sets. Furthermore, EDM had a superior mean subjective rating score (4.80 ± 0.42) for visualization of GGNs compared to other reconstructed images (all P < 0.001), although the model-based iterative reconstruction had superior subjective scores of overall image quality. For pure GGNs, the measured diameter and volume did not significantly differ among different reconstructions (both P > 0.05). EDM derived from DLSCT enabled improved image quality and lesion conspicuity for the evaluation of lung subsolid nodules compared to conventional iterative reconstruction algorithms and VMIs.

Accelerated High-Resolution Deep Learning Reconstruction Turbo Spin Echo MRI of the Knee at 7 T.

The aim of this study was to compare the image quality of 7 T turbo spin echo (TSE) knee images acquired with varying factors of parallel-imaging acceleration reconstructed with deep learning (DL)-based and conventional algorithms. This was a prospective single-center study. Twenty-three healthy volunteers underwent 7 T knee magnetic resonance imaging. Two-, 3-, and 4-fold accelerated high-resolution fat-signal-suppressing proton density (PD-fs) and T1-weighted coronal 2D TSE acquisitions with an encoded voxel volume of 0.31 × 0.31 × 1.5 mm3 were acquired. Each set of raw data was reconstructed with a DL-based and a conventional Generalized Autocalibrating Partially Parallel Acquisition (GRAPPA) algorithm. Three readers rated image contrast, sharpness, artifacts, noise, and overall quality. Friedman analysis of variance and the Wilcoxon signed rank test were used for comparison of image quality criteria. The mean age of the participants was 32.0 ± 8.1 years (15 male, 8 female). Acquisition times at 4-fold acceleration were 4 minutes 15 seconds (PD-fs, Supplemental Video is available at http://links.lww.com/RLI/A938) and 3 minutes 9 seconds (T1, Supplemental Video available at http://links.lww.com/RLI/A939). At 4-fold acceleration, image contrast, sharpness, noise, and overall quality of images reconstructed with the DL-based algorithm were significantly better rated than the corresponding GRAPPA reconstructions (P < 0.001). Four-fold accelerated DL-reconstructed images scored significantly better than 2- to 3-fold GRAPPA-reconstructed images with regards to image contrast, sharpness, noise, and overall quality (P ≤ 0.031). Image contrast of PD-fs images at 2-fold acceleration (P = 0.087), image noise of T1-weighted images at 2-fold acceleration (P = 0.180), and image artifacts for both sequences at 2- and 3-fold acceleration (P ≥ 0.102) of GRAPPA reconstructions were not rated differently than those of 4-fold accelerated DL-reconstructed images. Furthermore, no significant difference was observed for all image quality measures among 2-fold, 3-fold, and 4-fold accelerated DL reconstructions (P ≥ 0.082). This study explored the technical potential of DL-based image reconstruction in accelerated 2D TSE acquisitions of the knee at 7 T. DL reconstruction significantly improved a variety of image quality measures of high-resolution TSE images acquired with a 4-fold parallel-imaging acceleration compared with a conventional reconstruction algorithm.

Color image restoration by filtering methods: a review

Digital images are corrupted with noise, and image denoising is an important step in image processing modules. In this review, the latest developments in filtering methods for color image restoration are analyzed. These algorithms are compared in terms of objective image quality measures and divided into major classes, such as spatial domain, switching and wavelet filtering methods. These classes are based on the particular methodology used in image denoising algorithms and further subdivided to show their classification in terms of noise models utilized, application style, and stages the filters applied in images. In particular, we present a review of filtering methods in color image denoising, published over the past two decades. Our classification and succinct descriptions of color image restoration by these mathematical filtering techniques and their characterizations can help choose the appropriate ones for various downstream image processing tasks.

Assessment of cone-beam CT technical image quality indicators and radiation dose for optimal STL model used in visual surgical planning.

The aim of this study was to identify cone-beam computed tomography (CBCT) protocols that offer an optimal balance between effective dose (ED) and 3D model for orthognathic virtual surgery planning, using CT as a reference, and to assess whether such protocols can be defined based on technical image quality metrics. Eleven CBCT (VISO G7, Planmeca Oy, Helsinki, Finland) scan protocols were selected out of 32 candidate protocols, based on ED and technical image quality measurements. Next, an anthropomorphic RANDO SK150 phantom was scanned using these 11 CBCT protocols and 2 CT scanners for bone quantity assessments. The resulting DICOM (Digital Imaging and Communications in Medicine) files were converted into Standard Tessellation Language (STL) models that were used for bone volume and area measurements in the predefined orbital region to assess the validity of each CBCT protocol for virtual surgical planning. The highest CBCT bone volume and area of the STL models were obtained using normal dose protocol (F2) and ultra-low dose protocol (J13), which resulted in 48% and 96% of the mean STL bone volume and 48% and 95% of the bone area measured on CT scanners, respectively. The normal dose CBCT protocol "F2" offered optimal bone area and volume balance for STL. The optimal CBCT protocol can be defined using contrast-to-noise ratio and modulation transfer function values that were similar to those of the reference CT scanners'. CBCT scanners with selected protocols can offer a viable alternative to CT scanners for acquiring STL models for virtual surgical planning at a lower effective dose.

Spatial domain noise removal filtering for low-resolution digital images

&lt;p&gt;n this research work, six different filters are applied on a low resolution 8 b/pixel gray-scale images, which operate on small sub-images (windows of 3×3 to 11×11 pixels). The enhanced images are used to compare the efficiency of the different six filters using the peak signal to noise ratio (PSNR) image quality measure. Noise peak elimination filter (PSNR)=36.63) outperforms others, such as median filter (PSNR=36.61), while corruption estimation (PSNR=36.03) significantly cuts processing time by only processing the corrupted pixels while maintaining image details. Mean filter (PSNR=34.05) is sensitive to outliers, which cause the image's sharpness and fine features to be lost. By avoiding averaging across edges, bimodal-averaging filter (PSNR=35.30), which improves on the mean filter, chooses the mean of the biggest population. The median-mean filtering (PSNR=36.32), which combines median and mean filters and determines the output pixel by averaging the median and some nearby pixels, is another improvement above averaging.&lt;/p&gt;

47‐2: Invited Paper: Modulated Analog Driving and Evaluation of Image Quality

Consumer demand for higher resolution, higher frame rate, and deeper color displays is driving the data rates of the connected video systems to the limits of affordability. At current high‐end displays, like 4K gaming at 360 Hz frame rate or 8K TV's at &gt;120 Hz frame rate, conventional data transports require parallel data‐pairs plus equalization circuits to deliver the required video data. These accommodations increase cost, power and complexity.HYPHY's revolutionary ‘modulated analog' (also referred to as “sampled Analog”) video transport is the solution to this impasse, demonstrably delivering 10X more video payload per “wire‐hertz” than traditional methods, delivering equal or better image quality.To emphasize the advantages of this method, HYPHY has developed an image and video quality measurement method that demonstrates the image quality of Modulated Analog to be at least on par with traditional display driving.