A method to extract image noise level from patient images in CT

A novel method for developing contrast-detail curves from clinical patient images based on statistical low-contrast detectability

Purpose. To develop a method to extract statistical low-contrast detectability (LCD) and contrast-detail (C-D) curves from clinical patient images.Method. We used the region of air surrounding the patient as an alternative for a homogeneous region within a patient. A simple graphical user interface (GUI) was created to set the initial configuration for region of interest (ROI), ROI size, and minimum detectable contrast (MDC). The process was started by segmenting the air surrounding the patient with a threshold between -980 HU (Hounsfield units) and -1024 HU to get an air mask. The mask was trimmed using the patient center coordinates to avoid distortion from the patient table. It was used to automatically place square ROIs of a predetermined size. The mean pixel values in HU within each ROI were calculated, and the standard deviation (SD) from all the means was obtained. The MDC for a particular target size was generated by multiplying the SD by 3.29. A C-D curve was obtained by iterating this process for the other ROI sizes. This method was applied to the homogeneous area from the uniformity module of an ACR CT phantom to find the correlation between the parameters inside and outside the phantom, for 30 thoracic, 26 abdominal, and 23 head images.Results. The phantom images showed a significant linear correlation between the LCDs obtained from outside and inside the phantom, with R2values of 0.67 and 0.99 for variations in tube currents and tube voltages. This indicated that the air region outside the phantom can act as a surrogate for the homogenous region inside the phantom to obtain the LCD and C-D curves.Conclusion. The C-D curves obtained from outside the ACR CT phantom show a strong linear correlation with those from inside the phantom. The proposed method can also be used to extract the LCD from patient images by using the region of air outside as a surrogate for a region inside the patient.

An automated technique for global noise level measurement in CT image with a conjunction of image gradient

Automated assessment of noise level in clinical computed tomography (CT) images is a crucial technique for evaluating and ensuring the quality of these images. There are various factors that can impact CT image noise, such as statistical noise, electronic noise, structure noise, texture noise, artifact noise, etc. In this study, a method was developed to measure the global noise index (GNI) in clinical CT scans due to the fluctuation of x-ray quanta. Initially, a noise map is generated by sliding a 10 × 10 pixel for calculating Hounsfield unit (HU) standard deviation and the noise map is further combined with the gradient magnitude map. By employing Boolean operation, pixels with high gradients are excluded from the noise histogram generated with the noise map. By comparing the shape of the noise histogram from this method with Christianson's tissue-type global noise measurement algorithm, it was observed that the noise histogram computed in anthropomorphic phantoms had a similar shape with a close GNI value. In patient CT images, excluding the HU deviation due the structure change demonstrated to have consistent GNI values across the entire CT scan range with high heterogeneous tissue compared to the GNI values using Christianson's tissue-type method. The proposed GNI was evaluated in phantom scans and was found to be capable of comparing scan protocols between different scanners. The variation of GNI when using different reconstruction kernels in clinical CT images demonstrated a similar relationship between noise level and kernel sharpness as observed in uniform phantom: sharper kernel resulted in noisier images. This indicated that GNI was a suitable index for estimating the noise level in clinical CT images with either a smooth or grainy appearance. The study's results suggested that the algorithm can be effectively utilized to screen the noise level for a better CT image quality control.

Improving the Safety of Computed Tomography Through Automated Quality Measurement: A Radiologist Reader Study of Radiation Dose, Image Noise, and Image Quality

OBJECTIVES

The Centers for Medicare and Medicaid Services funded the development of a computed tomography (CT) quality measure for use in pay-for-performance programs, which balances automated assessments of radiation dose with image quality to incentivize dose reduction without compromising the diagnostic utility of the tests. However, no existing quantitative method for assessing CT image quality has been validated against radiologists' image quality assessments on a large number of CT examinations. Thus to develop an automated measure of image quality, we tested the relationship between radiologists' subjective ratings of image quality with measurements of radiation dose and image noise.

MATERIALS AND METHODS

Board-certified, posttraining, clinically active radiologists rated the image quality of 200 diagnostic CT examinations from a set of 734, representing 14 CT categories. Examinations with significant distractions, motion, or artifact were excluded. Radiologists rated diagnostic image quality as excellent, adequate, marginally acceptable, or poor; the latter 2 were considered unacceptable for rendering diagnoses. We quantified the relationship between ratings and image noise and radiation dose, by category, by analyzing the odds of an acceptable rating per standard deviation (SD) increase in noise or geometric SD (gSD) in dose.

RESULTS

One hundred twenty-five radiologists contributed 24,800 ratings. Most (89%) were acceptable. The odds of an examination being rated acceptable statistically significantly increased per gSD increase in dose and decreased per SD increase in noise for most categories, including routine dose head, chest, and abdomen-pelvis, which together comprise 60% of examinations performed in routine practice. For routine dose abdomen-pelvis, the most common category, each gSD increase in dose raised the odds of an acceptable rating (2.33; 95% confidence interval, 1.98-3.24), whereas each SD increase in noise decreased the odds (0.90; 0.79-0.99). For only 2 CT categories, high-dose head and neck/cervical spine, neither dose nor noise was associated with ratings.

CONCLUSIONS

Radiation dose and image noise correlate with radiologists' image quality assessments for most CT categories, making them suitable as automated metrics in quality programs incentivizing reduction of excessive radiation doses.

Ultra-High-Resolution CT of the Head and Neck with Deep Learning Reconstruction-Assessment of Image Quality and Radiation Exposure and Intraindividual Comparison with Normal-Resolution CT

OBJECTIVES

To assess the benefits of ultra-high-resolution CT (UHR-CT) with deep learning-based image reconstruction engine (AiCE) regarding image quality and radiation dose and intraindividually compare it to normal-resolution CT (NR-CT).

METHODS

Forty consecutive patients with head and neck UHR-CT with AiCE for diagnosed head and neck malignancies and available prior NR-CT of a different scanner were retrospectively evaluated. Two readers evaluated subjective image quality using a 5-point Likert scale regarding image noise, image sharpness, artifacts, diagnostic acceptability, and assessability of various anatomic regions. For reproducibility, inter-reader agreement was analyzed. Furthermore, signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and slope of the gray-value transition between different tissues were calculated. Radiation dose was evaluated by comparing CTDIvol, DLP, and mean effective dose values.

RESULTS

UHR-CT with AiCE reconstruction led to significant improvement in subjective (image noise and diagnostic acceptability: p < 0.000; ICC ≥ 0.91) and objective image quality (SNR: p < 0.000; CNR: p < 0.025) at significantly lower radiation doses (NR-CT 2.03 ± 0.14 mSv; UHR-CT 1.45 ± 0.11 mSv; p < 0.0001) compared to NR-CT.

CONCLUSIONS

Compared to NR-CT, UHR-CT combined with AiCE provides superior image quality at a markedly lower radiation dose. With improved soft tissue assessment and potentially improved tumor detection, UHR-CT may add further value to the role of CT in the assessment of head and neck pathologies.

A New Algorithm for Automatically Calculating Noise, Spatial Resolution, and Contrast Image Quality Metrics: Proof-of-Concept and Agreement With Subjective Scores in Phantom and Clinical Abdominal CT

OBJECTIVES

The aims of this study were to develop a proof-of-concept computer algorithm to automatically determine noise, spatial resolution, and contrast-related image quality (IQ) metrics in abdominal portal venous phase computed tomography (CT) imaging and to assess agreement between resulting objective IQ metrics and subjective radiologist IQ ratings.

MATERIALS AND METHODS

An algorithm was developed to calculate noise, spatial resolution, and contrast IQ parameters. The algorithm was subsequently used on 2 datasets of anthropomorphic phantom CT scans, acquired on 2 different scanners (n = 57 each), and on 1 dataset of patient abdominal CT scans (n = 510). These datasets include a range of high to low IQ: in the phantom dataset, this was achieved through varying scanner settings (tube voltage, tube current, reconstruction algorithm); in the patient dataset, lower IQ images were obtained by reconstructing 30 consecutive portal venous phase scans as if they had been acquired at lower mAs. Five noise, 1 spatial, and 13 contrast parameters were computed for the phantom datasets; for the patient dataset, 5 noise, 1 spatial, and 18 contrast parameters were computed. Subjective IQ rating was done using a 5-point Likert scale: 2 radiologists rated a single phantom dataset each, and another 2 radiologists rated the patient dataset in consensus. General agreement between IQ metrics and subjective IQ scores was assessed using Pearson correlation analysis. Likert scores were grouped into 2 categories, "insufficient" (scores 1-2) and "sufficient" (scores 3-5), and differences in computed IQ metrics between these categories were assessed using the Mann-Whitney U test.

RESULTS

The algorithm was able to automatically calculate all IQ metrics for 100% of the included scans. Significant correlations with subjective radiologist ratings were found for 4 of 5 noise (R2 range = 0.55-0.70), 1 of 1 spatial resolution (R2 = 0.21 and 0.26), and 10 of 13 contrast (R2 range = 0.11-0.73) parameters in the phantom datasets and for 4 of 5 noise (R2 range = 0.019-0.096), 1 of 1 spatial resolution (R2 = 0.11), and 16 of 18 contrast (R2 range = 0.008-0.116) parameters in the patient dataset. Computed metrics that significantly differed between "insufficient" and "sufficient" categories were 4 of 5 noise, 1 of 1 spatial resolution, 9 and 10 of 13 contrast parameters for phantom the datasets and 3 of 5 noise, 1 of 1 spatial resolution, and 10 of 18 contrast parameters for the patient dataset.

CONCLUSION

The developed algorithm was able to successfully calculate objective noise, spatial resolution, and contrast IQ metrics of both phantom and clinical abdominal CT scans. Furthermore, multiple calculated IQ metrics of all 3 categories were in agreement with subjective radiologist IQ ratings and significantly differed between "insufficient" and "sufficient" IQ scans. These results demonstrate the feasibility and potential of algorithm-determined objective IQ. Such an algorithm should be applicable to any scan and may help in optimization and quality control through automatic IQ assessment in daily clinical practice.

Tin filter compared to low kV protocols - optimizing sinonasal imaging in computed tomography

OBJECTIVES

Paranasal sinus imaging due to chronic inflammatory disease is one of the most common examinations in head and neck radiology with CT imaging considered the current gold standard. In this phantom study we analyzed different low dose CT protocols in terms of image quality, radiation exposure and subjective evaluation in order to establish an optimized scanning protocol.

METHODS

In a phantom study, an Alderson phantom was scanned using 12 protocols between 70-120 kV and 25-200 mAs with and without tin filtration. For all datasets, iterative reconstruction was used. Data were objectively evaluated (image noise, (dose-weighted) contrast-to-noise ratio) and for subjective evaluation an online survey using a Likert scale was performed to reach a large group of clinically experienced reader (n = 62). The protocol was considered diagnostically insufficient if the median score was 4 and above and if more than 10% of raters scored 4 and above on the Likert scale. For an interreader agreement an ICC was calculated. To compare clinical value in relation to the applied dose and the objective image parameters, we calculated a figure of merit (FOM) and ranked the protocols accordingly.

RESULTS

There was an overall moderate agreement between the 62 readers for the 12 examined CT protocols. In this phantom study, protocols with 100 kV with spectral shaping and 50-100 mAs obtained the best results for its combination of dose, image quality and clinical information value for diagnosing sinusitis (FOM 1st- 2nd place) with the 70 kV and 50 mAs as a good alternative as well (Sinusitis: FOM shared 2nd). For preoperative planning, where a higher dose is necessary, 100 kV with spectral shaping and 100 mAs achieved the overall best results (FOM 1st place) with 70 kV and 50 mAs ranking 4th.

CONCLUSION

100-kV protocols with spectral shaping or low kV protocols (70 kV) with a similarly low dose showed the best figure of merit for imaging sinonasal disease and preoperative planning. With modern scanner technology available, spectral shaping or low KV protocols should be used for sinusitis imaging.

Low-Dose CT Imaging of the Pelvis in Follow-up Examinations-Significant Dose Reduction and Impact of Tin Filtration: Evaluation by Phantom Studies and First Systematic Retrospective Patient Analyses

OBJECTIVES

Low-dose (LD) computed tomography (CT) is still rarely used in musculoskeletal (MSK) radiology. This study evaluates the potentials of LD CT for follow-up pelvic imaging with special focus on tin filtration (Sn) technology for normal and obese patients with and without metal implants.

MATERIALS AND METHODS

In a phantom study, 5 different LD and normal-dose (ND) CT protocols with and without tin filtration were tested using a normal and an obese phantom. Iterative reconstruction (IR) and filtered back projection (FBP) were used for CT image reconstruction. In a subsequent retrospective patient study, ND CT images of 45 patients were compared with follow-up tin-filtered LD CT images with a 90% dose reduction. Sixty-four percent of patients contained metal implants at the follow-up examination. Computed tomography images were objectively (image noise, contrast-to-noise ratio [CNR], dose-normalized contrast-to-noise ratio [CNRD]) and subjectively, using a 6-point Likert score, evaluated. In addition, the figure of merit was calculated. For group comparisons, paired t tests, Wilcoxon signed rank test, analysis of variance, or Kruskal-Wallis tests were used, where applicable.

RESULTS

The LD Sn protocol with 67% dose reduction resulted in equal values in qualitative (Likert score) and quantitative image analysis (image noise) compared with the ND protocol in the phantom study. For follow-up examinations, dose could be reduced up to 90% by using Sn LD CT scans without impairment in the clinical study. However, metal implants resulted in a mild impairment of Sn LD as well as ND CT images. Cancellous bone ( P < 0.001) was assessed worse and cortical bone ( P = 0.063) equally in Sn LD CT images compared with ND CT images. Figure of merit values were significant ( P ≤ 0.02) lower and hence better in Sn LD as in ND protocols. Obese patients benefited in particular from tin filtration in LD MSK imaging in terms of image noise and CNR ( P ≤ 0.05).

CONCLUSIONS

Low-dose CT scans with tin filtration allow maximum dose reduction while maintaining high image quality for certain clinical purposes, for example, follow-up examinations, especially metal implant position, material loosening, and consolidation controls. Overweight patients benefit particularly from tin filter technology. Although metal implants decrease image quality in ND as well as in Sn LD CT images, this is not a relevant limitation for assessability.

Evaluation of Load-To-Strength Ratios in Metastatic Vertebrae and Comparison With Age- and Sex-Matched Healthy Individuals

Vertebrae containing osteolytic and osteosclerotic bone metastases undergo pathologic vertebral fracture (PVF) when the lesioned vertebrae fail to carry daily loads. We hypothesize that task-specific spinal loading patterns amplify the risk of PVF, with a higher degree of risk in osteolytic than in osteosclerotic vertebrae. To test this hypothesis, we obtained clinical CT images of 11 cadaveric spines with bone metastases, estimated the individual vertebral strength from the CT data, and created spine-specific musculoskeletal models from the CT data. We established a musculoskeletal model for each spine to compute vertebral loading for natural standing, natural standing + weights, forward flexion + weights, and lateral bending + weights and derived the individual vertebral load-to-strength ratio (LSR). For each activity, we compared the metastatic spines' predicted LSRs with the normative LSRs generated from a population-based sample of 250 men and women of comparable ages. Bone metastases classification significantly affected the CT-estimated vertebral strength (Kruskal-Wallis, p < 0.0001). Post-test analysis showed that the estimated vertebral strength of osteosclerotic and mixed metastases vertebrae was significantly higher than that of osteolytic vertebrae (p = 0.0016 and p = 0.0003) or vertebrae without radiographic evidence of bone metastasis (p = 0.0010 and p = 0.0003). Compared with the median (50%) LSRs of the normative dataset, osteolytic vertebrae had higher median (50%) LSRs under natural standing (p = 0.0375), natural standing + weights (p = 0.0118), and lateral bending + weights (p = 0.0111). Surprisingly, vertebrae showing minimal radiographic evidence of bone metastasis presented significantly higher median (50%) LSRs under natural standing (p < 0.0001) and lateral bending + weights (p = 0.0009) than the normative dataset. Osteosclerotic vertebrae had lower median (50%) LSRs under natural standing (p < 0.0001), natural standing + weights (p = 0.0005), forward flexion + weights (p < 0.0001), and lateral bending + weights (p = 0.0002), a trend shared by vertebrae with mixed lesions. This study is the first to apply musculoskeletal modeling to estimate individual vertebral loading in pathologic spines and highlights the role of task-specific loading in augmenting PVF risk associated with specific bone metastatic types. Our finding of high LSRs in vertebrae without radiologically observed bone metastasis highlights that patients with metastatic spine disease could be at an increased risk of vertebral fractures even at levels where lesions have not been identified radiologically.

Fully automated image quality evaluation on patient CT: Multi-vendor and multi-reconstruction study

While the recent advancements of computed tomography (CT) technology have contributed in reducing radiation dose and image noise, an objective evaluation of image quality in patient scans has not yet been established. In this study, we present a patient-specific CT image quality evaluation method that includes fully automated measurements of noise level, structure sharpness, and alteration of structure. This study used the CT images of 120 patients from four different CT scanners reconstructed with three types of algorithm: filtered back projection (FBP), vendor-specific iterative reconstruction (IR), and a vendor-agnostic deep learning model (DLM, ClariCT.AI, ClariPi Inc.). The structure coherence feature (SCF) was used to divide an image into the homogeneous (RH) and structure edge (RS) regions, which in turn were used to localize the regions of interests (ROIs) for subsequent analysis of image quality indices. The noise level was calculated by averaging the standard deviations from five randomly selected ROIs on RH, and the mean SCFs on RS was used to estimate the structure sharpness. The structure alteration was defined by the standard deviation ratio between RS and RH on the subtraction image between FBP and IR or DLM, in which lower structure alterations indicate successful noise reduction without degradation of structure details. The estimated structure sharpness showed a high correlation of 0.793 with manually measured edge slopes. Compared to FBP, IR and DLM showed 34.38% and 51.30% noise reduction, 2.87% and 0.59% lower structure sharpness, and 2.20% and -12.03% structure alteration, respectively, on an average. DLM showed statistically superior performance to IR in all three image quality metrics. This study is expected to contribute to enhance the CT protocol optimization process by allowing a high throughput and quantitative image quality evaluation during the introduction or adjustment of lower-dose CT protocol into routine practice.

Automatic head computed tomography image noise quantification with deep learning

PURPOSE

Computed tomography (CT) image noise is usually determined by standard deviation (SD) of pixel values from uniform image regions. This study investigates how deep learning (DL) could be applied in head CT image noise estimation.

METHODS

Two approaches were investigated for noise image estimation of a single acquisition image: direct noise image estimation using supervised DnCNN convolutional neural network (CNN) architecture, and subtraction of a denoised image estimated with denoising UNet-CNN experimented with supervised and unsupervised noise2noise training approaches. Noise was assessed with local SD maps using 3D- and 2D-CNN architectures. Anthropomorphic phantom CT image dataset (N = 9 scans, 3 repetitions) was used for DL-model comparisons. Mean square error (MSE) and mean absolute percentage errors (MAPE) of SD values were determined using the SD values of subtraction images as ground truth. Open-source clinical head CT low-dose dataset (N_train = 37, N_test = 10 subjects) were used to demonstrate DL applicability in noise estimation from manually labeled uniform regions and in automated noise and contrast assessment.

RESULTS

The direct SD estimation using 3D-CNN was the most accurate assessment method when comparing in phantom dataset (MAPE = 15.5%, MSE = 6.3HU). Unsupervised noise2noise approach provided only slightly inferior results (MAPE = 20.2%, MSE = 13.7HU). 2DCNN and unsupervised UNet models provided the smallest MSE on clinical labeled uniform regions.

CONCLUSIONS

DL-based clinical image assessment is feasible and provides acceptable accuracy as compared to true image noise. Noise2noise approach may be feasible in clinical use where no ground truth data is available. Noise estimation combined with tissue segmentation may enable more comprehensive image quality characterization.

Medical physics 3.0: A renewed model for practicing medical physics in clinical imaging

Inspired by the principles of Medical Physics 3.0, this paper frames a new model of clinical physics practice, anchored to clinical realities, clinical priorities, and advanced physics. The model is based on the conviction that the physicist is vested with the expertise and the responsibility to ensure each patient gets the optimum imaging exam towards the best clinical outcome. Key expectations and activities are encapsulated into 12 areas: scientific perspective, quality and safety assurance, regulatory compliance, technology assessment, use optimization, performance monitoring, technology acquisition, technology commissioning, vendor cooperation, translational practice, research consultancy, and technology education. The paper further highlights key challenges to effective clinical physics practice of increased scope of competency, balancing rigor and relevance, managing metrological surrogates of quality and safety, and integrating principle- and data-informed approaches. Mindful to practically mitigate these challenges, clinical imaging physics can play an essential role to enable evidence-based imaging care.

Medical Physics 3.0 and Its Relevance to Radiology

Medical Physics 3.0 is a grassroots movement within the medical physics community to re-envision a new commitment and engagement of physics in the care process. In this model, a physicist, present in either the clinic or the laboratory, is to practice physics for the direct benefit of the patient, either immediately in the clinical setting or eventually through research. As a demonstration of this shift of focus, this article offers a framework by which medical physics can better serve the rigor and quality of radiological practice. This is accomplished through (1) patient-centered metrics of imaging dose and quality that are generalizable across and relatable to varying tasks; (2) prospectively prescribing protocols and dose based on needed quality; and (3) retrospectively monitoring and targeting dose and quality across practice. The article also suggests a model for this opportunity in which the current "consultant" physicist is vested with the responsibility to fully contribute her essential expertise in the care of individual patients.

Development, validation, and relevance of in vivo low-contrast task transfer function to estimate detectability in clinical CT images

PURPOSE

The current state-of-the-art calculation of detectability index (d') is largely phantom-based, with the latest being based on a hybrid phantom noise power spectrum (NPS) combined with patient-specific noise magnitude and high-contrast air-skin interface. The purpose of this study was to develop and assess the use of fully patient-specific measurements of noise and low-contrast resolution, derived entirely from patient images on d'.

METHODS

This study developed a d' calculation that is patient- and task-specific, employing newly developed algorithms for estimating patient-specific NPS and low-contrast task transfer function (TTF). The TTF estimation methodology used a trained regression support vector machine (SVM) to estimate a fitted form of the TTF given a variance-normalized estimate of the NPS (referred to as the TTF_NPS ). The regression SVM was trained and tested using five-fold cross-validation on 192 scans (4 dose levels x 6 reconstruction kernels x 4 repeats) of a phantom with low-contrast polyethylene insert and reconstructed with filtered backprojection and iterative reconstructions across 12 clinically relevant kernels (FBP: B20f, B31f, B45f; SAFIRE: I26f, I31f, J45f with strengths: 2, 3, 5). To test the low-contrast TTF estimation method, the estimated TTF_NPS measurements were compared to (1) TTF measurements from the air-phantom interface (referred to as the TTF_air , representing the most patient-specific clinical alternative) and (2) TTF measurements from the edge of the low-contrast polyethylene insert (referred to as the TTF_poly ), which represented the gold standard of low-contrast TTF measurement. Patient-specific NPS, patient-specific noise magnitude, and patient-specific low-contrast TTF were further combined with a reference task function to calculate a d' (according to a non-prewhitening matched filter model) across 1120 lesions previously evaluated in 2AFC human observer detection of liver lesions. The resulting values were compared to the observer results using a generalized linear mixed-effects statistical model. The correlations between the model and observer results were also compared with previously reported values (using a hybrid method with phantom-derived NPS and TTF_air ).

RESULTS

The TTF_NPS more accurately represented resolution across the considered reconstruction settings, compared with the TTF_air . The out-of-fold predictions of the TTF_NPS had statistically better root-mean-square error concordance (p < 0.05, one-tailed Wilcoxon rank-sum test) to gold standard than the TTF_air (the alternative, measured from the air-phantom interface). Detectability indices informed by purely patient-specific NPS and TTF were strongly correlated with 2AFC outcomes (p < 0.05). R² between human detection accuracy and model-predicted detection accuracy were shown to be greater for those measured with patient-specific d' than for the hybrid d' but failed to rise to the level of statistical significance (p ≥ 0.05, bootstrap resampled corrected paired Student's t-test).

CONCLUSIONS

The results suggest that fully patient-specific characterization of image quality based on in vivo NPS and low-contrast TTF offer advantages over hybrid methods. The results in terms of d' favorably relate to observer detection of liver lesions. The method can potentially be integrated into an automated image quality tracking system to assess image quality across a computed tomography clinical operation without needing phantom scans.

Assessment of the global noise algorithm for automatic noise measurement in head CT examinations

PURPOSE

The global noise (GN) algorithm has been previously introduced as a method for automatic noise measurement in clinical CT images. The accuracy of the GN algorithm has been assessed in abdomen CT examinations, but not in any other body part until now. This work assesses the GN algorithm accuracy in automatic noise measurement in head CT examinations.

METHODS

A publicly available image dataset of 99 head CT examinations was used to evaluate the accuracy of the GN algorithm in comparison to reference noise values. Reference noise values were acquired using a manual noise measurement procedure. The procedure used a consistent instruction protocol and multiple observers to mitigate the influence of intra- and interobserver variation, resulting in precise reference values. Optimal GN algorithm parameter values were determined. The GN algorithm accuracy and the corresponding statistical confidence interval were determined. The GN measurements were compared across the six different scan protocols used in this dataset. The correlation of GN to patient head size was also assessed using a linear regression model, and the CT scanner's X-ray beam quality was inferred from the model fit parameters.

RESULTS

Across all head CT examinations in the dataset, the range of reference noise was 2.9-10.2 HU. A precision of ±0.33 HU was achieved in the reference noise measurements. After optimization, the GN algorithm had a RMS error 0.34 HU corresponding to a percent RMS error of 6.6%. The GN algorithm had a bias of +3.9%. Statistically significant differences in GN were detected in 11 out of the 15 different pairs of scan protocols. The GN measurements were correlated with head size with a statistically significant regression slope parameter (p < 10-7 ). The CT scanner X-ray beam quality estimated from the slope parameter was 3.5 cm water HVL (2.8-4.8 cm 95% CI).

CONCLUSION

The GN algorithm was validated for application in head CT examinations. The GN algorithm was accurate in comparison to reference manual measurement, with errors comparable to interobserver variation in manual measurement. The GN algorithm can detect noise differences in examinations performed on different scanner models or using different scan protocols. The trend in GN across patients of different head sizes closely follows that predicted by a physical model of X-ray attenuation.

Improved precision of noise estimation in CT with a volume-based approach

Assessment of image noise is a relevant issue in computed tomography (CT). Noise is routinely measured by the standard deviation of density values (Hounsfield units, HU) within a circular region of interest (ROI). We explored the effect of a spherical volume of interest (VOI) on noise measurements. Forty-nine chronic obstructive pulmonary disease patients underwent CT with clinical protocol (regular dose [RD], volumetric CT dose index [CTDIvol] 3.04 mGy, 64-slice unit), and ultra-low dose (ULD) protocol (median CTDIvol 0.38 mGy, dual-source unit). Noise was measured in 27 1-cm² ROIs and 27 0.75-cm³ VOIs inside the trachea. Median true noise was 21 HU (range 17-29) for RD-CT and 33 HU (26-39) for ULD-CT. The VOI approach resulted in a lower mean distance between limits of agreement compared to ROI: 5.9 versus 10.0 HU for RD-CT (-40%); 4.7 versus 9.9 HU for ULD-CT (-53%). Mean systematic bias barely changed: -1.6 versus -0.9HU for RD-CT; 0.0 to 0.4HU for ULD-CT. The average measurement time was 6.8 s (ROI) versus 9.7 (VOI), independent of dose level. For chest CT, measuring noise with a VOI-based instead of a ROI-based approach reduces variability by 40-53%, without a relevant effect on systematic bias and measurement time.

Development and validation of an automated methodology to assess perceptual in vivo noise texture in liver CT

Purpose: Developing, validating, and evaluating a method for measuring noise texture directly from patient liver CT images (i.e., in vivo). Approach: The method identifies target regions within patient scans that are least likely to have major contribution of patient anatomy, detrends them locally, and measures noise power spectrum (NPS) there using a previously phantom-validated technique targeting perceptual noise-non-anatomical fluctuations in the image that may interfere with the detection of focal lesions. Method development and validation used scanner-specific CT simulations of computational, anthropomorphic phantom (XCAT phantom, three phases of contrast-enhancement) with known ground truth of the NPS. Simulations were based on a clinical scanner (Definition Flash, Siemens) and clinically relevant settings (tube voltage of 120 kV at three dose levels). Images were reconstructed with filtered backprojection (kernel: B31, B41, and B50) and Sinogram Affirmed Iterative Reconstruction (kernel: I31, I41, and I50) using a manufacturer-specific reconstruction software (ReconCT, Siemens). All NPS measurements were made in the liver. Ground-truth NPS were taken as the sum of (1) a measurement in parenchymal regions of anatomy-subtracted (i.e., noise only) scans, and (2) a measurement in the same region of noise-free (pre-noise-insertion) images. To assess in vivo NPS performance, correlation of NPS average frequency (f avg ), was reported. Sensitivity of accuracy [root-mean-square-error (RMSE)] to number of pixels included in measurement was conducted via bootstrapped pixel-dropout. Sensitivity of NPS to dose and reconstruction kernel was assessed to confirm that ground truth NPS similarities were maintained in patient-specific measurements. Results: Pearson and Spearman correlation coefficients 0.97 and 0.96 forf avg indicated good correlation. Results suggested accurate NPS measurements (within 5% total RMSE) could be acquired with ∼ 10 6    pixels . Conclusions: Relationships of similar NPS due to reconstruction kernel and dose were preserved between gold standard and observed in vivo estimations. The NPS estimation method was further deployed on clinical cases to demonstrate the feasibility of clinical analysis.

A Benchmark for automatic noise measurement in clinical computed tomography

PURPOSE

Assessment of image quality directly in clinical image data is an important quality control objective as phantom-based testing does not fully represent image quality across patient variation. Computer algorithms for automatically measuring noise in clinical computed tomography (CT) images have been introduced, but the accuracy of these algorithms is unclear. This work benchmarks the accuracy of the global noise (GN) algorithm for automatic noise measurement in contrast-enhanced abdomen CT exams in comparison to precise reference noise measurements. The GN algorithm was further optimized compared to the previous report in the literature.

METHODS

Reference values of noise were established in a public image dataset of 82 contrast-enhanced abdomen CT exams. The reference noise values were obtained by manual regions-of-interest measurements of pixel standard deviation in the liver parenchyma according to an instruction protocol. Noise measurements taken by six observers were averaged together to improve reference noise statistical precision. The GN algorithm was used to automatically measure noise in each image set. The accuracy of the GN algorithm was determined in terms of RMS error compared to reference noise. The GN algorithm was optimized by conducting 1000 trials with random algorithm parameter values. The trial with the lowest RMS error was used to select optimum algorithm parameters.

RESULTS

The range of noise across CT image sets was 8.8-28.8 HU. Reference noise measurements were made with a precision of ±0.78 HU (95% confidence interval). The RMS error of automatic noise measurement was 0.93 HU (0.77-1.19 HU 95% confidence interval). The automatic noise measurements were equally accurate across image sets of varying noise magnitude. Optimum GN algorithm parameter values were: a kernel size of 7 pixels, and soft tissue lower and upper thresholds of 0 and 170 HU, respectively.

CONCLUSIONS

The performance of automatic noise measurement was benchmarked in a large clinical CT dataset. The study provides a framework for thorough validation of automatic clinical image quality measurement methods. The GN algorithm was optimized and validated for automatic measurement of soft-tissue noise in abdomen CT exams.

Development of a computational phantom for validation of automated noise measurement in CT images

The purpose of this study was to develop a computational phantom for validation of automatic noise calculations applied to all parts of the body, to investigate kernel size in determining noise, and to validate the accuracy of automatic noise calculation for several noise levels. The phantom consisted of objects with a very wide range of HU values, from -1000 to +950. The incremental value for each object was 10 HU. Each object had a size of 15 × 15 pixels separated by a distance of 5 pixels. There was no dominant homogeneous part in the phantom. The image of the phantom was then degraded to mimic the real image quality of CT by convolving it with a point spread function (PSF) and by addition of Gaussian noise. The magnitude of the Gaussian noises was varied (5, 10, 25, 50, 75 and 100 HUs), and they were considered as the ground truth noise (N_G). We also used a computational phantom with added actual noise from a CT scanner. The phantom was used to validate the automated noise measurement based on the average of the ten smallest standard deviations (SD) from the standard deviation map (SDM). Kernel sizes from 3 × 3 up to 27 × 27 pixels were examined in this study. A computational phantom for automated noise calculations validation has been successfully developed. It was found that the measured noise (N_M) was influenced by the kernel size. For kernels of 15 × 15 pixels or smaller, the N_Mvalue was much smaller than the N_G. For kernel sizes from 17 × 17 to 21 × 21 pixels, the N_Mvalue was about 90% of N_G. And for kernel sizes of 23 × 23 pixels and above, N_Mis greater than N_G. It was also found that even with small kernel sizes the relationship between N_Mand N_Gis linear with R²more than 0.995. Thus accurate noise levels can be automatically obtained even with small kernel sizes without any concern regarding the inhomogeneity of the object.

Dose Reduction in Dental CT: A Phantom Study With Special Focus on Tin Filter Technique

OBJECTIVE. The purpose of this study was to determine in a phantom the dose exposure of different dental 3D sectional imaging methods (CT and cone-beam CT [CBCT]) and different CT protocols. The aim was to establish optimal protocols with the lowest possible dose and diagnostically high image quality with special consideration given to tin prefiltration. MATERIALS AND METHODS. Dose was determined with thermoluminescence detectors at 20 different measuring points on an anthropomorphic phantom. Eight different CT protocols with and without tin filtering were compared with iterative reconstruction methods and a standard CBCT protocol. Objective and subjective image evaluations and a figure-of-merit analysis of the image data were performed by radiologists and maxillofacial surgeons. RESULTS. The determined dose-length products of the nine examinations were 5.0-111.9 mGy · cm with a calculated effective whole body dose of 20.7-505.9 μSv. Cone-beam CT was in the upper midfield with an effective dose of 229.3 μSv. On the basis of dose, objective image quality, and clinical evaluation results, tin filter protocols performed best. Protocols with higher doses were significantly less useful in the figure of merit comparison but because of their detailed bony representation are particularly necessary to answer certain questions about trauma and tumors. CONCLUSION. The use of tin filtering can reduce dose in dental CT examinations, compared with standard low-dose examinations, while maintaining good image quality. The dose performance is significantly inferior even to that of a cone-beam CT examination. High-dose protocols are necessary only for certain questions.

Validation of algorithmic CT image quality metrics with preferences of radiologists

PURPOSE

Automated assessment of perceptual image quality on clinical Computed Tomography (CT) data by computer algorithms has the potential to greatly facilitate data-driven monitoring and optimization of CT image acquisition protocols. The application of these techniques in clinical operation requires the knowledge of how the output of the computer algorithms corresponds to clinical expectations. This study addressed the need to validate algorithmic image quality measurements on clinical CT images with preferences of radiologists and determine the clinically acceptable range of algorithmic measurements for abdominal CT examinations.

MATERIALS AND METHODS

Algorithmic measurements of image quality metrics (organ HU, noise magnitude, and clarity) were performed on a clinical CT image dataset with supplemental measures of noise power spectrum from phantom images using techniques developed previously. The algorithmic measurements were compared to clinical expectations of image quality in an observer study with seven radiologists. Sets of CT liver images were selected from the dataset where images in the same set varied in terms of one metric at a time. These sets of images were shown via a web interface to one observer at a time. First, the observer rank ordered the CT images in a set according to his/her preference for the varying metric. The observer then selected his/her preferred acceptable range of the metric within the ranked images. The agreement between algorithmic and observer rankings of image quality were investigated and the clinically acceptable image quality in terms of algorithmic measurements were determined.

RESULTS

The overall rank-order agreements between algorithmic and observer assessments were 0.90, 0.98, and 1.00 for noise magnitude, liver parenchyma HU, and clarity, respectively. The results indicate a strong agreement between the algorithmic and observer assessments of image quality. Clinically acceptable thresholds (median) of algorithmic metric values were (17.8, 32.6) HU for noise magnitude, (92.1, 131.9) for liver parenchyma HU, and (0.47, 0.52) for clarity.

CONCLUSIONS

The observer study results indicated that these algorithms can robustly assess the perceptual quality of clinical CT images in an automated fashion. Clinically acceptable ranges of algorithmic measurements were determined. The correspondence of these image quality assessment algorithms to clinical expectations paves the way toward establishing diagnostic reference levels in terms of clinically acceptable perceptual image quality and data-driven optimization of CT image acquisition protocols.