Different Aspects Of Image Quality Research Articles

Deep convolutional-neural-network-based metal artifact reduction for CT-guided interventional oncology procedures (MARIO).

Computed tomography (CT) is routinely used to guide cryoablation procedures. Notably, CT-guidance provides 3D localization of cryoprobes and can be used to delineate frozen tissue during ablation. However, metal-induced artifacts from ablation probes can make accurate probe placement challenging and degrade the ice ball conspicuity, which in combination could lead to undertreatment of potentially curable lesions. In this work, we propose an image-based neural network (CNN) model for metal artifact reduction for CT-guided interventional procedures. An image domain metal artifact simulation framework was developed and validated for deep-learning-based metal artifact reduction for interventional oncology (MARIO). CT scans were acquired for 19 different cryoablation probe configurations. The probe configurations varied in the number of probes and the relative orientations. A combination of intensity thresholding and masking based on maximum intensity projections (MIPs) was used to segment both the probes only and probes + artifact in each phantom image. Each of the probe and probe + artifact images were then inserted into 19 unique patient exams, in the image domain, to simulate metal artifact appearance for CT-guided interventional oncology procedures. The resulting 361 pairs of simulated image volumes were partitioned into disjoint training and test datasets of 304 and 57 volumes, respectively. From the training partition, 116600 image patches with a shape of 128×128×5 pixels were randomly extracted to be used for training data. The input images consisted of a superposition of the patient and probe + artifact images. The target images consisted of a superposition of the patient and probe only images. This dataset was used to optimize a U-Net type model. The trained model was then applied to 50 independent, previously unseen CT images obtained during renal cryoablations. Three board-certified radiologists with experience in CT-guided ablations performed a blinded review of the MARIO images. A total of 100 images (50 original, 50 MARIO processed) were assessed across different aspects of image quality on a 4-point likert-type item. Statistical analyses were performed using Wilcoxon signed-rank test for paired samples. Reader scores were significantly higher for MARIO processed images compared to the original images across all metrics (all p<0.001). The average scores of the overall image quality, iceball conspicuity, overall metal artifact, needle tip visualization, target region confidence, and worst metal artifact, needle tip visualization, iceball conspicuity, and target region confidence improved by 34.91%, 36.29%, 39.94%, 34.17%, 35.13%, and 45.70%, respectively. The proposed method of image-based metal artifact simulation can be used to train a MARIO algorithm to effectively reduce probe-related metal artifacts in CT-guided cryoablation procedures.

Multi-reader evaluation of different image quality scoring systems in prostate MRI

ObjectivesTo evaluate different image quality scoring systems in the assessment of factors limiting diagnostic accuracy of prostate MRI. MethodsThis retrospective IRB-approved study included 281 patients undergoing prostate MRI prior to biopsy. Four readers (2 experienced, 2 novice) independently reviewed all MRI examinations (n = 295) and assigned scores for subjective image quality (1–5; 1:poor, 5:excellent), the PI-QUAL and the PSHS scoring system. The original PI-RADS scores were extracted from the report and transperineal template saturation biopsy served as histopathological reference. ResultsInter-reader agreement was found to be good, with PSHS showing highest agreement (kappa: 0.65). The PSHS scoring system performed well assessing the influence of image quality on sensitivity of MR for clinically-significant cancer for the experienced readers using a PI-RADS score cut-off ≥ 3/≥4, as did the PI-QUAL scoring system with a PI-RADS cut-off ≥ 4. For the less experienced radiologist, this was true for PSHS (clinically-significant and all cancers) and PI-QUAL scores (clinically-significant cancers) for a PI-RADS score ≥ 3. PSHS scores were positively associated with the detection of clinically-significant cancer based on a PI-RADS cut-off ≥ 4, OR 1.86 (95 % CI 1.22–2.82), and had the highest Somers’ D. ConclusionsThe PSHS scoring system performed well in assessing the effect of image quality on detection rates, as did the PI-QUAL system. Since both systems focus on different aspects of image quality, their incorporation into prostate MRI reports could further enhance standardization and allow for a reliable assessment of image quality as a potential confounder in prostate MRI.

Reproducibility and feasibility of optic nerve diffusion MRI techniques: single-shot echo-planar imaging (EPI), readout-segmented EPI, and reduced field-of-view diffusion-weighted imaging

BackgroundDiffusion-weighted imaging (DWI) is an essential technique for optic nerve diseases. However, the image quality of optic nerve DWI is decreased by the distortions and artifacts associated with conventional techniques. In order to establish this method as a critical tool in optic nerve diseases, reproducibility and feasibility of new technical and conventional approaches of DWI need to be systematically investigated.MethodsDWIs were acquired using ss-EPI, readout-segmented EPI (rs-EPI) DWI, and reduced field-of-view (rFOV) DWI. 26 volunteers (mean age 31.2 years) underwent repeated MRI examinations in order to assess scan–rescan reproducibility and accuracy. The apparent diffusion coefficient (ADC) values (three ROIs were measured on each side) were determined to evaluate the reproducibility of each sequence and the differences between the three techniques. To quantify the geometric distortion artifacts, the length of optic nerve and the maximum angle of optic nerve were defined and compared to T2-weighted imaging. In addition, two readers evaluated four different aspects of image quality on 5-point Likert scales.Resultsrs-EPI DWI (ICCs: 0.916, 0.797 and 0.781) and rFOV DWI (ICCs: 0.850, 0.595 and 0.750) showed higher reproducibility (ICCs: ROI1, ROI2 and ROI3) of mean ADC value in all three ROIs than ss-EPI DWI (ICCs: 0.810, 0.442 and 0.379). The quantitative analysis of geometric distortion yielded a higher agreement of both rs-EPI DWI and rFOV DWI with T2-weighted imaging than ss-EPI. rs-EPI DWI (2.38 ± 0.90) and rFOV DWI (2.46 ± 0.58) were superior to ss-EPI DWI (1.58 ± 0.64) with respect to overall image quality and other aspects of image quality, each with P < 0.05. The mean ADC values of rFOV DWI were significantly lower than those of rs-EPI DWI and ss-EPI DWI in all three ROIs (P < 0.001).ConclusionsBoth rs-EPI DWI and rFOV-EPI DWI are suitable techniques for the assessment of diffusion restriction and provide significantly improved image quality compared with ss-EPI DWI. For methods using the same acquisition time, rFOV DWI is superior to ss-EPI DWI, while rs-EPI showed an overall superiority, although this technique took 47% longer to perform.

Improved Readout-Segmented Echo-Planner Diffusion-Weighted Magnetic Resonance Imaging of Nasopharyngeal Carcinoma Using Simultaneous Multislice Acquisitions at 3 T.

PurposeThis study systematically compared the images from readout-segmented echo-planar diffusion-weighted imaging (RESOLVE-DWI [RS-DWI]) and simultaneous multislice accelerated RESOLVE-DWI (SMS-RS-DWI) in patients with nasopharyngeal carcinoma (NPC) in qualitative and quantitative aspects.MethodForty-four patients with NPC were included. The RS-DWI and prototypic SMS-RS-DWI sequences were performed on all patients. Images were qualitatively evaluated by 4 independent radiologists using a 5-point Likert scale. For quantitative evaluation, the maximum and minimum diameters and the maximum tumor areas were determined for both DWI sequences and compared with the T2-weighted imaging (T2WI) to evaluate image distortions. The apparent diffusion coefficient was measured in the slice with the maximum tumor profile.ResultsThe SMS-RS-DWI was superior to RS-DWI with respect to overall image quality (3.40 ± 0.53 vs 2.71 ± 0.48, P < 0.0001) and tumor edge sharpness (3.29 ± 0.65 vs 2.64 ± 0.47, P < 0.0001). Susceptibility artifacts were significantly less severe in SMS-RS-DWI than in RS-DWI (0.85 ± 0.57 vs 1.36 ± 0.57, P < 0.0001). There was no significant overestimation or underestimation of the tumor geometry using the SMS-RS-DWI or RS-DWI compared with T2WI. The quantitative analysis showed a slightly higher agreement for SMS-RS-DWI with T2WI than RS-DWI for maximum diameter, minimum diameter, and maximum tumor area. The apparent diffusion coefficient values showed no significant differences between the 2 DWI techniques (P > 0.05).ConclusionsAt 3 T, SMS-RS-DWI is a useful technique for diagnosing NPC. It substantially improves different aspects of image quality by providing higher spatial resolution and fewer susceptibility artifacts with more extensive anatomic coverage compared with RS-DWI.

Image quality is resilient against tube voltage variations in post-mortem skeletal radiography with a digital flat-panel detector

In recent phantom studies low-contrast detectability was shown to be independent from variations in tube voltage in digital radiography (DR) systems. To investigate the transferability to a clinical setting, the lower extremities of human cadavers were exposed at constant detector doses with different tube voltages in a certain range, as proposed in the phantom studies. Three radiologists independently graded different aspects of image quality (IQ) in a comparative analysis. The grades show no correlation between IQ and kV, which means that the readers were not able to recognize a significant IQ difference at different kV. Signal-to-noise and contrast-to-noise ratios showed no significant differences in IQ despite the kV-setting variations. These findings were observed from a limited kV range setting. Higher kV-settings resulted in lowest patient exposure at constant IQ. These results confirm the potential of DR-systems to contribute to standardization of examination protocols comparable to computed tomography. This may prevent the trend to overexpose. Further investigations in other body regions and other DR-systems are encouraged to determine transferability.

Inclusion of quasi-vertex views in a brain-dedicated multi-pinhole SPECT system for improved imaging performance

With brain-dedicated multi-detector systems employing pinhole apertures the usage of detectors facing the top of the patient’s head (i.e. quasi-vertex (QV) views) can provide the advantage of additional viewing from close to the brain for improved detector coverage. In this paper, we report the results of simulation and reconstruction studies to investigate the impact of the QV views on the imaging performance of AdaptiSPECT-C, a brain-dedicated stationary SPECT system under development. In this design, both primary and scatter photons from regions located inferior to the brain can contribute to SPECT projections acquired by the QV views, and thus degrade AdaptiSPECT-C imaging performance. In this work, we determined the proportion, origin, and nature (i.e. primary, scatter, and multiple-scatter) of counts emitted from structures within the head and throughout the body contributing to projections from the different AdaptiSPECT-C detector rings, as well as from a true vertex view detector. We simulated phantoms used to assess different aspects of image quality (i.e. uniform activity concentration sphere, and Derenzo), as well as anthropomorphic phantoms with different count levels emulating clinical 123I activity distributions (i.e. DaTscan and perfusion). We determined that attenuation and scatter in the patient’s body greatly diminish the probability of the photons emitted outside the volume of interest reaching to detectors and being recorded within the 15% photopeak energy window. In addition, we demonstrated that the inclusion of the residual of such counts in the system acquisition does not degrade visual interpretation or quantitative analysis. The addition of the QV detectors improves volumetric sensitivity, angular sampling, and spatial resolution leading to significant enhancement in image quality, especially in the striato-thalamic and superior regions of the brain. Besides, the use of QV detectors improves the recovery of clinically relevant metrics such as the striatal binding ratio and mean activity in selected cerebral structures. Our findings proving the usefulness of the QV ring for brain imaging with 123I agents can be generalized to other commonly used SPECT imaging agents labelled with isotopes, such as 99mTc and likely 111In.

A Framework for Objective Evaluation of Single Image De-Hazing Techniques

Real-world environment, where images are acquired with digital camera, may be subject to sever climatic conditions such as haze that may drastically reduce the quality performance of sophisticated computer vision algorithms used for various tasks, e.g., tracking, detection, classification etc. Even though several single image de-hazing techniques have been recently proposed with many deep-learning approaches among them, a general statistical framework that would permit an objective performance evaluation has not been independently introduced yet. In this manuscript, certain performance metrics that emphasize different aspects of image quality, output ranges and polarity, are identified and combined into a single performance indicator derived in an unbiased manner. A general methodology is thus introduced, as a framework for objective performance evaluation of current and future dehazing tasks, through an extensive comparison of 15 single image de-hazing techniques over a vast range of image data sets. The proposed unified framework shows several advantages in evaluating diverse and perceptually meaningful image features but also in elucidating future directions for improvement in image dehazing tasks.

Wavelet-based reconstruction of dynamic susceptibility MR-perfusion: a new method to visualize hypervascular brain tumors.

Parameter maps based on wavelet-transform post-processing of dynamic perfusion data offer an innovative way of visualizing blood vessels in a fully automated, user-independent way. The aims of this study were (i) a proof of concept regarding wavelet-based analysis of dynamic susceptibility contrast (DSC) MRI data and (ii) to demonstrate advantages of wavelet-based measures compared to standard cerebral blood volume (CBV) maps in patients with the initial diagnosis of glioblastoma (GBM). Consecutive 3-T DSC MRI datasets of 46 subjects with GBM (mean age 63.0 ± 13.1years, 28 m) were retrospectively included in this feasibility study. Vessel-specific wavelet magnetic resonance perfusion (wavelet-MRP) maps were calculated using the wavelet transform (Paul wavelet, order 1) of each voxel time course. Five different aspects of image quality and tumor delineation were each qualitatively rated on a 5-point Likert scale. Quantitative analysis included image contrast and contrast-to-noise ratio. Vessel-specific wavelet-MRP maps could be calculated within a mean time of 2:27min. Wavelet-MRP achieved higher scores compared to CBV in all qualitative ratings: tumor depiction (4.02 vs. 2.33), contrast enhancement (3.93 vs. 2.23), central necrosis (3.86 vs. 2.40), morphologic correlation (3.87 vs. 2.24), and overall impression (4.00 vs. 2.41); all p < .001. Quantitative image analysis showed a better image contrast and higher contrast-to-noise ratios for wavelet-MRP compared to conventional perfusion maps (all p < .001). wavelet-MRP is a fast and fully automated post-processing technique that yields reproducible perfusion maps with a clearer vascular depiction of GBM compared to standard CBV maps. • Wavelet-MRP offers high-contrast perfusion maps with a clear delineation of focal perfusion alterations. • Both image contrast and visual image quality were beneficial for wavelet-MRP compared to standard perfusion maps like CBV. • Wavelet-MRP can be automatically calculated from existing dynamic susceptibility contrast (DSC) perfusion data.

Diffusion-weighted MRI of the prostate: advantages of Zoomed EPI with parallel-transmit-accelerated 2D-selective excitation imaging.

The purpose of our study was to evaluate the use of 2D-selective, parallel-transmit excitation magnetic resonance imaging (MRI) for diffusion-weighted echo-planar imaging (pTX-EPI) of the prostate, and to compare it to conventional, single-shot EPI (c-EPI). The MRI examinations of 35 patients were evaluated in this prospective study. PTX-EPI was performed with a TX-acceleration factor of 1.7 and a field of view (FOV) of 150 × 90 mm(2), whereas c-EPI used a full FOV of 380 × 297 mm(2). Two readers evaluated three different aspects of image quality on 5-point Likert scales. To quantify distortion artefacts, maximum diameters and prostate volume were determined for both techniques and compared to T2-weighted imaging. The zoomed pTX-EPI was superior to c-EPI with respect to overall image quality (3.39 ± 0.62 vs 2.45 ± 0.67) and anatomic differentiability (3.29 ± 0.65 vs 2.41 ± 0.65), each with p < 0.0001. Artefacts were significantly less severe in pTX-EPI (0.93 ± 0.73 vs 1.49 ± 1.08), p < 0.001. The quantitative analysis yielded a higher agreement of pTX-EPI with T2-weighted imaging than c-EPI with respect to coronal (ICCs: 0.95 vs 0.93) and sagittal (0.86 vs 0.73) diameters as well as prostate volume (0.94 vs 0.92). Apparent diffusion coefficient (ADC) values did not differ significantly between the two techniques (p > 0.05). Zoomed pTX-EPI leads to substantial improvements in diffusion-weighted imaging (DWI) of the prostate with respect to different aspects of image quality and severity of artefacts. Recent technical developments in MRI allow the use of accelerated, spatially-selective excitation (parallel-transmit, pTX). pTX can be used for zoomed echo-planar prostate imaging (pTX-EPI). pTX-EPI improves different aspects of image quality in prostate MRI. Distortion artefacts are reduced by the use of pTX-EPI in prostate MRI. Further studies should aim at assessing the diagnostic accuracy of pTX-EPI.

Parallel-transmit-accelerated spatially-selective excitation mri for reduced-fov diffusion-weighted-imaging of the pancreas

ObjectivesTo find out whether the use of accelerated 2D-selective parallel-transmit excitation MRI for diffusion-weighted EPI (pTX-EPI) offers advantages over conventional single-shot EPI (c-EPI) with respect to different aspects of image quality in the MRI of the pancreas. Materials and methodsThe MRI examinations of 33 consecutive patients were evaluated in this prospective and IRB-approved study. PTX-EPI was performed with a reduced (zoomed) FOV of 230×118mm2. The 2D-RF pulse of pTX-EPI was accelerated, i.e. shortened by a factor of 1.7 (pTX-acceleration factor). C-EPI used a full-FOV of 380×285mm2. In a qualitative analysis, two experienced readers evaluated 3 different aspects of image quality on 3- to 5-point Likert scales. Additionally, apparent diffusion coefficients (ADCs) were determined in both c-EPI and pTX-EPI in normal-appearing pancreatic tissue using regions of interests (ROIs). Mean ADC values and standard deviations were compared between the two techniques. ResultsThe reduced-FOV pTX-EPI was superior to c-EPI with respect to overall image quality (p<0.0001) and identifiability of the pancreatic ducts (p<0.01). Artifacts were significantly less severe in pTX-EPI (p<0.01). The mean ADC values of c-EPI (1.29±0.19×10−3mm2/s) and pTX-EPI (1.27±0.17×10−3mm2/s) did not differ significantly between the two techniques (p=0.44). The variation within the ROIs as measured by the standard deviation was significantly lower in pTX-EPI (0.095×10−3mm2/s) than in c-EPI (0.135×10−3mm2/s), p<0.05. ConclusionsPTX-accelerated EPI with spatially-selective excitation and reduced FOV leads to substantial improvements in DWI of the pancreas with respect to different aspects of image quality without significantly influencing the ADC values.

8-bit or 11-bit monochrome displays—which image is preferred by the radiologist?

Vendors of medical displays promise a better diagnostic performance using 10- or 11-bit instead of 8-bit monochrome displays. We measured the gain of "Just Noticeable Difference" (JND) steps using high grayscale resolutions and evaluated the preference of radiologists regarding different aspects of image quality. The amount of JND steps was measured on a display using 8-, 10- and 11-bit Look Up Tables (LUT). Radiological images were presented simultaneously using an 8-bit and an 11-bit LUT, radiologists where asked to determine the better image presentation regarding quality parameters like sharpness, contrast and detectability of details. The 10-bit technology realized more than twice the number of JND steps compared to an 8-bit LUT, an 11-bit LUT presented only few additional JND steps. The radiologists did not attest the higher grayscale resolution a better image quality, they regarded the 8-bit technology to show a better sharpness and contrast, although this had no impact on the detectability of details. Our measurements confirmed that the higher grayscale resolution results in a more complete visualization of image information. But radiologists partially judged this as a lack of sharpness and contrast and generally preferred the 8-bit display.

Comparison of conventional B-scan, tissue harmonic imaging, compound imaging and tissue harmonic compound imaging in neck lesion characterisation

In recent years, further technical developments of ultrasound scanning techniques, such as tissue harmonic imaging (THI) and compound imaging (CI), have become available and promise considerable improvement in image quality. No comparative assessments have yet been made of their systematic use in the head and neck. We studied 313 lesions of the head and neck detected on ultrasound scanning. Ultrasound images were obtained using a state-of-the-art scanning system. Two experts evaluated the images obtained for each lesion with conventional B-scan mode (BSCAN), THI, CI, and tissue harmonic compound imaging (THICI) with respect to four different aspects of image quality. Largely concordant results were found for each of the parameters studied: overall image quality, tissue contrast, lesion conspicuity, and internal structure. Evaluations of CI and THICI were frequently ranked higher (p < 0.001) than BSCAN and THI (p < 0.001). Images obtained in BSCAN mode often had better scores than images in THI mode alone (p < 0.001). Comparison of CI and THICI showed no statistically significant differences for any of the parameters studied. After a learning period, compound imaging methods improve image quality of the soft tissues of neck and may be included in the routine settings of ultrasound systems.

MO‐A‐ValB‐01: Tradeoffs in Image Quality and Radiation Dose for CT

In CT scanning, image quality has many components and is influenced by many technical parameters. While image quality has always been a concern for the physics community, clinically acceptable image quality has become even more of an issue as strategies to reduce radiation dose — to all patients, but especially to pediatric patients— has become a focus in many radiology practices.The purpose of this presentation will be to first describe several of the components of CT image quality — noise, slice thickness (Z‐axis resolution), low contrast resolution and high contrast resolution— as well as radiation dose and to describe how each of these may be affected by technical parameter selection. This presentation will pay particular attention to the tradeoffs that exist between different aspects of image quality, especially when the reduction of radiation dose is one of the objectives.The presentation will then explore several mechanisms that can be used to reduce radiation dose in CT exams and the implications for the diagnostic image quality of the exam. Specifically, the implications of varying the tube current*time product (mAs), pitch or tablespeed (or for axial imaging, the table increment), slice thickness, beam energy (kVp), patient (or phantom) size and dose reduction options (such as tube current modulation) will be described for both radiation dose and diagnostic image quality. Finally, this presentation will emphasize that the tradeoffs between radiation dose and image quality are clinical‐task dependent; that is, the goals of the clinically indicated exam dictate what aspect of image quality may be emphasized for that exam (low contrast resolution or high contrast spatial resolution, etc.) and this will have implications for the amount of radiation dose reduction that is acceptable. This will be illustrated with examples from selected diagnostic imaging exams.Educational Objectives:1. Understand key components of image quality in CT scanning as well as reinforce CT radiation dose concepts.2. Understand the impact that technical parameter selection has on the various aspects of image quality and radiation dose.3. Examine the tradeoffs between various aspects of image quality and radiation dose.4. Examine the impact of these tradeoffs on a few clinical imaging protocols and illustrate the task‐dependence of image quality requirements.

Wavelet-based image fusion and quality assessment

Recent developments in satellite and sensor technologies have provided high-resolution satellite images. Image fusion techniques can improve the quality, and increase the application of these data. This paper addresses two issues in image fusion (a) the image fusion method and (b) corresponding quality assessment.Firstly, a multi-band wavelet-based image fusion method is presented, which is a further development of the two-band wavelet transformation. This fusion method is then applied to a case study to demonstrate its performance in image fusion.Secondly, quality assessment for fused images is discussed. The objectives of image fusion include enhancing the visibility of the image and improving the spatial resolution and the spectral information of the original images. For assessing quality of an image after fusion, we define the aspects to be assessed initially. These include, for instance, spatial and spectral resolution, quantity of information, visibility, contrast, or details of features of interest. Quality assessment is application dependant; different applications may require different aspects of image quality. Based on this analysis, a set of qualities is classified and analyzed. These sets of qualities include (a) average grey value, for representing intensity of an image, (b) standard deviation, information entropy, profile intensity curve for assessing details of fused images, and (c) bias and correlation coefficient for measuring distortion between the original image and fused image in terms of spectral information.