Search

Objectives: The purpose of this study was to determine the influence of dose reduction on a commercially available lung cancer prediction convolutional neuronal network (LCP-CNN)., Methods: CT scans from a cohort provided by the local lung cancer center (n = 218) with confirmed pulmonary malignancies and their corresponding reduced dose simulations (25% and 5% dose) were subjected to the LCP-CNN. The resulting LCP scores (scale 1-10, increasing malignancy risk) and the proportion of correctly classified nodules were compared. The cohort was divided into a low-, medium-, and high-risk group based on the respective LCP scores; shifts between the groups were studied to evaluate the potential impact on nodule management. Two different malignancy risk score thresholds were analyzed: a higher threshold of ≥ 9 ("rule-in" approach) and a lower threshold of > 4 ("rule-out" approach)., Results: In total, 169 patients with 196 nodules could be included (mean age ± SD, 64.5 ± 9.2 year; 49% females). Mean LCP scores for original, 25% and 5% dose levels were 8.5 ± 1.7, 8.4 ± 1.7 (p > 0.05 vs. original dose) and 8.2 ± 1.9 (p < 0.05 vs. original dose), respectively. The proportion of correctly classified nodules with the "rule-in" approach decreased with simulated dose reduction from 58.2 to 56.1% (p = 0.34) and to 52.0% for the respective dose levels (p = 0.01). For the "rule-out" approach the respective values were 95.9%, 96.4%, and 94.4% (p = 0.12). When reducing the original dose to 25%/5%, eight/twenty-two nodules shifted to a lower, five/seven nodules to a higher malignancy risk group., Conclusion: CT dose reduction may affect the analyzed LCP-CNN regarding the classification of pulmonary malignancies and potentially alter pulmonary nodule management., Clinical Relevance Statement: Utilization of a "rule-out" approach with a lower malignancy risk threshold prevents underestimation of the nodule malignancy risk for the analyzed software, especially in high-risk cohorts., Key Points: • LCP-CNN may be affected by CT image parameters such as noise resulting from low-dose CT acquisitions. • CT dose reduction can alter pulmonary nodule management recommendations by affecting the outcome of the LCP-CNN. • Utilization of a lower malignancy risk threshold prevents underestimation of pulmonary malignancies in high-risk cohorts., (© 2023. The Author(s).)

Abstract: Photon-counting computed tomography (PCCT) offers better high-resolution and noise performance than energy integrating detector (EID) CT. In this work, we compared both technologies for imaging of the temporal bone and skull base. A clinical PCCT system and 3 clinical EID CT scanners were used to image the American College of Radiology image quality phantom using a clinical imaging protocol with matched CTDI vol (CT dose index-volume) of 25 mGy. Images were used to characterize the image quality of each system across a series of high-resolution reconstruction options. Noise was calculated from the noise power spectrum, whereas resolution was quantified with a bone insert by calculating a task transfer function. Images of an anthropomorphic skull phantom and 2 patient cases were examined for visualization of small anatomical structures. Across measured conditions, PCCT had a comparable or smaller average noise magnitude (120 Hounsfield units [HU]) to the EID systems (144-326 HU). Photon-counting CT also had comparable resolution (task transfer function f25 : 1.60 mm -1 ) to the EID systems (1.34-1.77 mm -1 ). Imaging results supported quantitative findings as PCCT more clearly showed the 12-lp/cm bars from the fourth section of the American College of Radiology phantom and better represented the vestibular aqueduct and oval and round windows when compared with the EID scanners. A clinical PCCT system was able to image the temporal bone and skull base with improved spatial resolution and lower noise than clinical EID CT systems at matched dose., Competing Interests: E.S. lists relationships with the following entities unrelated to the present publication: GE, Bracco, Imalogix, 12Sigma, Metis Health Analytics, Cambridge University Press, and Wiley and Sons. J.B.S. benefits from a license agreement with Sun Nuclear Corporation and is an adviser for METIS Health Analytics. Both relationships are unrelated to this work. The other authors declare no conflict of interest., (Copyright © 2023 Wolters Kluwer Health, Inc. All rights reserved.)

Rationale and Objectives: The accuracy of measured radiomics features is affected by CT imaging protocols. This study aims to ascertain if applying bias corrections can improve the classification performance of the radiomics features., Materials and Methods: A cohort of 144 Non-Small Cell Lung Cancer patient CT images was used to calculate radiomics features for use in predictive models of patient pathological stage. Simulation models of the tumors, matched to patient lesion qualities of size, contrast, and degree of spiculation, were used to both create and assess protocol-specific correction factors. The usefulness of correction was first assessed by applying the corrections to simulated lesion phantoms with known properties using a corrected paired Student's t-test. The sensitivity of radiomics features to correction factors was assessed by applying a library of possible theoretical correction factors to the uncorrected radiomics from the patient data. The data were then used to assess the effect of the correction on prediction performance (AUC) from a logistic regression radiomics model across the patient cases., Results: The correction factors were shown to reduce the bias of radiomics features, caused by protocols, provided that the correction factors were derived from lesion models with similar properties. The sensitivity of the radiomics features to changes due to protocol effects was on average 89% among all features. The corrections applied to patient data resulted in a small increase of 0.0074 in AUC that was not statistically significant (p=0.60)., Conclusion: Protocol-specific correction factors can be applied to radiomics studies to control for biases introduced by different imaging protocols. The correction factors should ideally be lesion-specific, derived using lesion models that echo patient lesion characteristics in terms of size, contrast, and degree of spiculation. Small corrections in the 10% range offers only a small improvement in the predictability of radiomics., (Copyright © 2021 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.)

Rationale and Objectives: The 3-fold purpose of this study was to (1) develop a method to relate measured differences in radiomics features in different computed tomography (CT) scans to one another and to true feature differences; (2) quantify minimum detectable change in radiomics features based on measured radiomics features from pairs of synthesized CT images acquired under variable CT scan settings, and (3) ascertain and inform the recommendations of the Quantitative Imaging Biomarkers Alliance (QIBA) for nodule volumetry., Materials and Methods: Images of anthropomorphic lung nodule models were simulated using resolution and noise properties for 297 unique imaging conditions. Nineteen morphology features were calculated from both the segmentation masks derived from the imaged nodules and from ground truth nodules. Analysis was performed to calculate minimum detectable difference of radiomics features as a function of imaging protocols in comparison to QIBA guidelines., Results: The minimum detectable differences ranged from 1% to 175% depending on the specific feature and set of imaging protocols. The results showed that QIBA protocol recommendations result in improved minimum detectable difference as compared to the range of possible protocols. The results showed that the minimum detectable differences may be improved from QIBA's current recommendation by further restricting the slice thickness requirement to be between 0.5 mm and 1 mm., Conclusion: Minimum detectable differences of radiomics features were quantified for lung nodules across a wide range of possible protocols. The results can be used prospectively to inform decision-making about imaging protocols to provide superior quantification of radiomics features., (Copyright © 2020 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.)

A novel routine dual-energy computed tomography (DECT) quality control (QC) program was developed to address the current deficiency of routine QC for this technology. The dual-energy quality control (DEQC) program features (1) a practical phantom with clinically relevant materials and concentrations, (2) a clinically relevant acquisition, reconstruction, and postprocessing protocol, and (3) a fully automated analysis software to extract quantitative data for database storage and trend analysis. The phantom, designed for easy set up for standalone or adjacent imaging next to the ACR phantom, was made in collaboration with an industry partner and informed by clinical needs to have four iodine inserts (0.5, 1, 2, and 5 mg/ml) and one calcium insert (100 mg/ml) equally spaced in a cylindrical water-equivalent background. The imaging protocol was based on a clinical DECT abdominal protocol capable of producing material specific concentration maps, virtual unenhanced images, and virtual monochromatic images. The QC automated analysis software uses open-source technologies which integrates well with our current automated CT QC database. The QC program was tested on a GE 750 HD scanner and two Siemens SOMATOM FLASH scanners over a 3-month period. The automated algorithm correctly identified the appropriate region of interest (ROI) locations and stores measured values in a database for monitoring and trend analysis. Slight variations in protocol settings were noted based on manufacturer. Overall, the project proved to provide a convenient and dependable clinical tool for routine oversight of DE CT imaging within the clinic., (© 2021 The Authors. Journal of Applied Clinical Medical Physics published by Wiley Periodicals, LLC on behalf of The American Association of Physicists in Medicine.)

Photon-counting CT detectors are the next step in advancing CT system development and will replace the current energy integrating detectors (EID) in CT systems in the near future. In this context, the performance of PCCT was compared to EID CT for three clinically relevant tasks: abdominal soft tissue imaging, where differentiating low contrast features is important; vascular imaging, where iodine detectability is critical; and, high-resolution skeletal and lung imaging. A multi-tiered phantom was imaged on an investigational clinical PCCT system (Siemens Healthineers) across different doses using three imaging modes: macro and ultra-high resolution (UHR) PCCT modes and EID CT. Images were reconstructed using filtered backprojection and soft tissue (B30f), vascular (B46f), or high-resolution (B70f; U70f for UHR) kernels. Noise power spectra, task transfer functions, and detectability index were evaluated. For a soft tissue task, PCCT modes showed comparable noise and resolution with improved contrast-to-noise ratio. For a vascular task, PCCT modes showed lower noise and improved iodine detectability. For a high resolution task, macro mode showed lower noise and comparable resolution while UHR mode showed higher noise but improved spatial resolution for both air and bone. PCCT offers competitive advantages to EID CT for clinical tasks.

Purpose: Phantoms are useful tools in diagnostic CT, but practical limitations reduce phantoms to being only a limited patient surrogate. Furthermore, a phantom with a single cross sectional area cannot be used to evaluate scanner performance in modern CT scanners that use dose reduction techniques such as automated tube current modulation (ATCM) and iterative reconstruction (IR) algorithms to adapt x-ray flux to patient size, reduce radiation dose, and achieve uniform image noise. A new multisized phantom (Mercury Phantom, MP) has been introduced, representing multiple diameters. This work aimed to ascertain if measurements from MP can predict radiation dose and image noise in clinical CT images to prospectively inform protocol design., Methods: The adult MP design included four different physical diameters (18.5, 23.0, 30.0, and 37.0 cm) representing a range of patient sizes. The study included 1457 examinations performed on two scanner models from two vendors, and two clinical protocols (abdominopelvic with and chest without contrast). Attenuating diameter, radiation dose, and noise magnitude (average pixel standard deviation in uniform image) was automatically estimated in patients and in the MP using a previously validated algorithm. An exponential fit of CTDI vol and noise as a function of size was applied to patients and MP data. Lastly, the fit equations from the phantom data were used to fit the patient data. In each patient distribution fit, the normalized root mean square error (nRMSE) values were calculated in the residuals' plots as a metric to indicate how well the phantom data can predict dose and noise in clinical operations as a function of size., Results: For dose across patient size distributions, the difference between nRMSE from patient fit and MP model data prediction ranged between 0.6% and 2.0% (mean 1.2%). For noise across patient size distributions, the nRMSE difference ranged between 0.1% and 4.7% (mean 1.4%)., Conclusions: The Mercury Phantom provided a close prediction of radiation dose and image noise in clinical patient images. By assessing dose and image quality in a phantom with multiple sizes, protocol parameters can be designed and optimized per patient size in a highly constrained setup to predict clinical scanner and ATCM system performance., (© 2020 American Association of Physicists in Medicine.)

OBJECTIVE. Diagnostic reference levels were developed as guidance for radiation dose in medical imaging and, by inference, diagnostic quality. The objective of this work was to expand the concept of diagnostic reference levels to explicitly include noise of CT examinations to simultaneously target both dose and quality through corresponding reference values. MATERIALS AND METHODS. The study consisted of 2851 adult CT examinations performed with scanners from two manufacturers and two clinical protocols: abdominopelvic CT with IV contrast administration and chest CT without IV contrast administration. An institutional informatics system was used to automatically extract protocol type, patient diameter, volume CT dose index, and noise magnitude from images. The data were divided into five reference patient size ranges. Noise reference level, noise reference range, dose reference level, and dose reference range were defined for each size range. RESULTS. The data exhibited strong dependence between dose and patient size, weak dependence between noise and patient size, and different trends for different manufacturers with differing strategies for tube current modulation. The results suggest size-based reference intervals and levels for noise and dose (e.g., noise reference level and noise reference range of 11.5-12.9 HU and 11.0-14.0 HU for chest CT and 10.1-12.1 HU and 9.4-13.7 HU for abdominopelvic CT examinations) that can be targeted to improve clinical performance consistency. CONCLUSION. New reference levels and ranges, which simultaneously consider image noise and radiation dose information across wide patient populations, were defined and determined for two clinical protocols. The methods of new quantitative constraints may provide unique and useful information about the goal of managing the variability of image quality and dose in clinical CT examinations.

Purpose: To create and validate a systematic observer performance platform for evaluation of simulated liver lesions at pediatric CT and to test this paradigm to measure the effect of radiation dose reduction on detection performance and reader confidence., Materials and Methods: Thirty normal pediatric (from patients aged 0-10 years) contrast material-enhanced, de-identified abdominal CT scans obtained from July 1, 2012, through July 1, 2016, were retrospectively collected from the clinical database. The study was exempt from institutional review board approval. Zero to three simulated, low-contrast liver lesions (≤6 mm) were digitally inserted by using software, and noise was added to simulate reductions in volume CT dose index (representing radiation dose estimation) of 25% and 50%. Pediatric, abdominal, and resident radiologists (three of each) reviewed 90 data sets in three sessions using an online interface, marking each lesion location and rating confidence (scale, 0-100). Statistical analysis was performed by using software., Results: Mixed-effects models revealed a significant decrease in detection sensitivity as radiation dose decreased ( P < .001). The mean confidence of the full-dose and 25% dose reduction examinations was significantly higher than that of the 50% dose reduction examinations ( P = .011 and .012, respectively) but not different from one another ( P = .866). Dose was not a significant predictor of time to complete each case, and subspecialty was not a significant predictor of sensitivity or false-positive results., Conclusion: Sensitivity for lesion detection significantly decreased as dose decreased; however, confidence did not change between the full-dose and 25% reduced-dose scans. This suggests that readers are unaware of this decrease in performance, which should be accounted for in clinical dose reduction efforts. Keywords: Abdomen/GI, CT, Liver, Observer Performance, Pediatrics, Perception Image© RSNA, 2019., Competing Interests: Disclosures of Conflicts of Interest: J.S.N. disclosed no relevant relationships. J.B.S. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: receives royalties (<$5000/year) from a license deal between Duke University and Sun Nuclear Gammex for the design of a CT image quality phantom. Other relationships: disclosed no relevant relationships. E.S. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: paid by Imalogix for board membership and consultancy; paid by for expert testimony (patent law); institution receives grants from Siemens and Bracco; institution receives royalties from GE, Imalogix, 12Sigma, SunNuclear, Cambridge University Press, and Wiley and Sons. Other relationships: disclosed no relevant relationships. T.R. disclosed no relevant relationships. L.N. disclosed no relevant relationships. A.E. disclosed no relevant relationships. B.Z. disclosed no relevant relationships. B.C.A. disclosed no relevant relationships. J.T.D. disclosed no relevant relationships. A.D. disclosed no relevant relationships. R.G. disclosed no relevant relationships. D.M. disclosed no relevant relationships. C.M.M. disclosed no relevant relationships. V.P. disclosed no relevant relationships. B.N.P. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: institution receives grant from Siemens; author receives grant from GE Healthcare; author receives payment for lectures from GE Healthcare. Other relationships: disclosed no relevant relationships. G.R.S. disclosed no relevant relationships. D.P.F. disclosed no relevant relationships., (2019 by the Radiological Society of North America, Inc.)

We aimed to design and fabricate synthetic lung nodules with patient-informed internal heterogeneity to assess the variability and accuracy of measured texture features in CT. To that end, 190 lung nodules from a publicly available database of chest CT images (Lung Image Database Consortium) were selected based on size ( > 3    mm ) and malignancy. The texture features of the nodules were used to train a statistical texture synthesis model based on clustered lumpy background. The model parameters were ascertained based on a genetic optimization of a Mahalanobis distance objective function. The resulting texture model defined internal heterogeneity within 24 anthropomorphic lesion models which were subsequently fabricated into physical phantoms using a multimaterial three-dimensional (3-D) printer. The 3-D-printed lesions were imbedded in an anthropomorphic chest phantom and imaged with a clinical scanner using different acquisition parameters including slice thickness, dose level, and reconstruction kernel. The imaged lesions were analyzed in terms of texture features to ascertain the impact of CT imaging on lesion texture quantification. The texture modeling method produced lesion models with low and stable Mahalanobis distance between real and synthetic textures. The virtual lesions were successfully printed into 3-D phantoms. The accuracy and variability of the measured features extracted from the CT images of the phantoms showed notable influence from the imaging acquisition parameters with contrast, energy, and texture entropy exhibiting most sensitivity in terms of accuracy, and contrast, dissimilarity, and texture entropy most variability. Thinner slice thicknesses yielded more accurate and edge reconstruction kernels more stable results. We conclude that printed textured models of lesions can be developed using a method that can target and minimize the mathematical distance between real and synthetic lesions. The synthetic lesions can be used as the basis to investigate how CT imaging conditions might affect radiomics features derived from CT images.

Purpose: The task-based assessment of image quality using model observers is increasingly used for the assessment of different imaging modalities. However, the performance computation of model observers needs standardization as well as a well-established trust in its implementation methodology and uncertainty estimation. The purpose of this work was to determine the degree of equivalence of the channelized Hotelling observer performance and uncertainty estimation using an intercomparison exercise., Materials and Methods: Image samples to estimate model observer performance for detection tasks were generated from two-dimensional CT image slices of a uniform water phantom. A common set of images was sent to participating laboratories to perform and document the following tasks: (a) estimate the detectability index of a well-defined CHO and its uncertainty in three conditions involving different sized targets all at the same dose, and (b) apply this CHO to an image set where ground truth was unknown to participants (lower image dose). In addition, and on an optional basis, we asked the participating laboratories to (c) estimate the performance of real human observers from a psychophysical experiment of their choice. Each of the 13 participating laboratories was confidentially assigned a participant number and image sets could be downloaded through a secure server. Results were distributed with each participant recognizable by its number and then each laboratory was able to modify their results with justification as model observer calculation are not yet a routine and potentially error prone., Results: Detectability index increased with signal size for all participants and was very consistent for 6 mm sized target while showing higher variability for 8 and 10 mm sized target. There was one order of magnitude between the lowest and the largest uncertainty estimation., Conclusions: This intercomparison helped define the state of the art of model observer performance computation and with thirteen participants, reflects openness and trust within the medical imaging community. The performance of a CHO with explicitly defined channels and a relatively large number of test images was consistently estimated by all participants. In contrast, the paper demonstrates that there is no agreement on estimating the variance of detectability in the training and testing setting., (© 2018 American Association of Physicists in Medicine.)

This study aims to characterize the effect of background tissue density and heterogeneity on the detection of irregular masses in breast tomosynthesis, while demonstrating the capability of the sophisticated tools that can be used in the design, implementation, and performance analysis of virtual clinical trials (VCTs). Twenty breast phantoms from the extended cardiac-torso (XCAT) family, generated based on dedicated breast computed tomography of human subjects, were used to extract a total of 2173 volumes of interest (VOIs) from simulated tomosynthesis images. Five different lesions, modeled after human subject tomosynthesis images, were embedded in the breasts and combined with the lesion absent condition yielded a total of [Formula: see text] VOIs. Effects of background tissue density and heterogeneity on the detection of the lesions were studied by implementing a composite hypothesis signal detection paradigm with location known exactly, lesion known exactly or statistically, and background known statistically. Using the area under the receiver operating characteristic curve, detection performance deteriorated as density was increased, yielding findings consistent with clinical studies. A human observer study was performed on a subset of the simulated tomosynthesis images, confirming the detection performance trends with respect to density and serving as a validation of the implemented detector. Performance of the implemented detector varied substantially across the 20 breasts. Furthermore, background tissue density and heterogeneity affected the log-likelihood ratio test statistic differently under lesion absent and lesion present conditions. Therefore, considering background tissue variability in tissue models can change the outcomes of a VCT and is hence of crucial importance. The XCAT breast phantoms have the potential to address this concern by offering realistic modeling of background tissue variability based on a wide range of human subjects, comprising various breast shapes, sizes, and densities.

Objective: Modern CT systems use surrogates of noise-noise index (NI) and quality reference effective tube current-time product (Q)-to infer the quality of images acquired using tube current modulation. This study aimed to determine the relationship between actual noise and these surrogates for two CT scanners from two different manufacturers., Materials and Methods: Two phantoms (adult and 1-year-old child) were imaged on two CT scanners (64 and 128 MDCT) using a clinical range of NI (6-22) and Q (30-300 mA). Each scan was performed twice, and noise was measured in the mediastinum, lung, and abdomen using an image subtraction technique. The effect on noise from changing other imaging parameters, such as beam collimation, pitch, peak kilovoltage, slice thickness, FOV, reconstruction kernel or algorithm, and patient age category (adult or pediatric), was investigated., Results: On the 64-MDCT scanner, noise increased linearly along with NI, with the slope affected by changing the anatomy of interest, peak kilovoltage, reconstruction algorithm, and convolution kernel. The noise-NI relationship was independent of phantom size, slice thickness, pitch, FOV, and beam width. On the 128-MDCT scanner, noise decreased nonlinearly along with increasing Q, slice thickness, and peak tube voltage. The noise-Q relationship also depended on anatomy of interest, phantom size, age selection, and reconstruction algorithm but was independent of pitch, FOV, and detector configuration., Conclusion: We established how noise changes with changing image quality indicators across a clinically relevant range of imaging parameters. This work can aid in optimizing protocols by targeting specific noise levels for different types of CT examinations.

Purpose: To quantitatively compare noise texture across computed tomography (CT) scanners from different manufacturers using the noise power spectrum (NPS)., Methods: The American College of Radiology CT accreditation phantom (Gammex 464, Gammex, Inc., Middleton, WI) was imaged on two scanners: Discovery CT 750HD (GE Healthcare, Waukesha, WI), and SOMATOM Definition Flash (Siemens Healthcare, Germany), using a consistent acquisition protocol (120 kVp, 0.625∕0.6 mm slice thickness, 250 mAs, and 22 cm field of view). Images were reconstructed using filtered backprojection and a wide selection of reconstruction kernels. For each image set, the 2D NPS were estimated from the uniform section of the phantom. The 2D spectra were normalized by their integral value, radially averaged, and filtered by the human visual response function. A systematic kernel-by-kernel comparison across manufacturers was performed by computing the root mean square difference (RMSD) and the peak frequency difference (PFD) between the NPS from different kernels. GE and Siemens kernels were compared and kernel pairs that minimized the RMSD and |PFD| were identified., Results: The RMSD (|PFD|) values between the NPS of GE and Siemens kernels varied from 0.01 mm(2) (0.002 mm(-1)) to 0.29 mm(2) (0.74 mm(-1)). The GE kernels "Soft," "Standard," "Chest," and "Lung" closely matched the Siemens kernels "B35f," "B43f," "B41f," and "B80f" (RMSD < 0.05 mm(2), |PFD| < 0.02 mm(-1), respectively). The GE "Bone," "Bone+," and "Edge" kernels all matched most closely with Siemens "B75f" kernel but with sizeable RMSD and |PFD| values up to 0.18 mm(2) and 0.41 mm(-1), respectively. These sizeable RMSD and |PFD| values corresponded to visually perceivable differences in the noise texture of the images., Conclusions: It is possible to use the NPS to quantitatively compare noise texture across CT systems. The degree to which similar texture across scanners could be achieved varies and is limited by the kernels available on each scanner.