Automated assessment of low contrast sensitivity for CT systems using a model observer

Hernandez-Giron, Irene; Geleijns, Jacob; Calzado, A.; Veldkamp, Wouter J. H.

doi:10.1118/1.3577757

Cited by 48 publications

(36 citation statements)

References 24 publications

(35 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…22 If one observer was inconsistent in his results between both sessions for a given condition, that scoring was ruled out. 16 The average human observer performance was obtained as the mean of the PC values that passed the intra-observer tests for each condition. Finally, PC curves, as a function of the object diameter, were obtained for each mA and either FBP or AIDR 3D.…”

Section: Image Acquisitionmentioning

confidence: 99%

“…26,27 For this 2-AFC experiment, fitting curves according to Equation (6) were applied for each mA and reconstruction set independently, for both the average human and model observer. 16 For the average human observer, the error bars related to the PC values, previously calculated, were used as weights in the fitting process based on a least-squares procedure. The range of the fitting curves runs from 0.5 (pure guessing) and 1.00 (certain detection).…”

Section: Psychometric Fits and Visibility Thresholdsmentioning

confidence: 99%

“…[12][13][14][15] In a previous work, an objective statistical method using a specific model observer [non-pre-whitening matched filter with an eye filter (NPWE)] was presented to investigate the influence of different CT acquisition parameters on LCDet. 16 The main goal of this work is to compare the model observer LCDet performance in CT images acquired at different dose levels with human observers. Images reconstructed with two algorithms (FBP and iterative) were used in this study.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Comparison between human and model observer performance in low-contrast detection tasks in CT images: application to images reconstructed with filtered back projection and iterative algorithms

Hernandez-Giron¹,

Calzado²,

Geleijns³

et al. 2014

BJR

View full text Add to dashboard Cite

RMS, Veldkamp WJH. Comparison between human and model observer performance in lowcontrast detection tasks in CT images: application to images reconstructed with filtered back projection and iterative algorithms. Br J Radiol 2014;87:20140014. FULL PAPERComparison between human and model observer performance in low-contrast detection tasks in CT images: application to images reconstructed with filtered back projection and iterative algorithms Objective: To compare low-contrast detectability (LCDet) performance between a model [non-pre-whitening matched filter with an eye filter (NPWE)] and human observers in CT images reconstructed with filtered back projection (FBP) and iterative [adaptive iterative dose reduction three-dimensional (AIDR 3D; Toshiba Medical Systems, Zoetermeer, Netherlands)] algorithms. Methods: Images of the Catphan® phantom (Phantom Laboratories, New York, NY) were acquired with Aquilion ONE™ 320-detector row CT (Toshiba Medical Systems, Tokyo, Japan) at five tube current levels (20-500 mA range) and reconstructed with FBP and AIDR 3D. Samples containing either low-contrast objects (diameters, 2-15 mm) or background were extracted and analysed by the NPWE model and four human observers in a two-alternative forced choice detection task study. Proportion correct (PC) values were obtained for each analysed object and used to compare human and model observer performances. An efficiency factor (h) was calculated to normalize NPWE to human results. Results: Human and NPWE model PC values (normalized by the efficiency, h 5 0.44) were highly correlated for the whole dose range. The Pearson's product-moment correlation coefficients (95% confidence interval) between human and NPWE were 0.984 (0.972-0.991) for AIDR 3D and 0.984 (0.971-0.991) for FBP, respectively. Bland-Altman plots based on PC results showed excellent agreement between human and NPWE [mean absolute difference 0.5 6 0.4%; range of differences (24.7%, 5.6%)]. Conclusion: The NPWE model observer can predict human performance in LCDet tasks in phantom CT images reconstructed with FBP and AIDR 3D algorithms at different dose levels. Advances in knowledge: Quantitative assessment of LCDet in CT can accurately be performed using software based on a model observer.

show abstract

Section: Image Acquisitionmentioning

confidence: 99%

Section: Psychometric Fits and Visibility Thresholdsmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Comparison between human and model observer performance in low-contrast detection tasks in CT images: application to images reconstructed with filtered back projection and iterative algorithms

Hernandez-Giron¹,

Calzado²,

Geleijns³

et al. 2014

BJR

View full text Add to dashboard Cite

show abstract

“…In the subslice module three groups of four inserts each are contained. Diameters range between 3 and 9 mm and contrast is fixed at 1.0% (4) . The Gammex ACR CT phantom (Gammex RMI, Middleton, WI) is made from Solid Water and has a 20 cm diameter and 16 cm length.…”

Section: Introductionmentioning

confidence: 99%

Development of a universal medical X‐ray imaging phantom prototype

Groenewald¹,

Groenewald

2016

J Applied Clin Med Phys

View full text Add to dashboard Cite

Diagnostic X‐ray imaging depends on the maintenance of image quality that allows for proper diagnosis of medical conditions. Maintenance of image quality requires quality assurance programs on the various X‐ray modalities, which consist of projection radiography (including mobile X‐ray units), fluoroscopy, mammography, and computed tomography (CT) scanning. Currently a variety of modality‐specific phantoms are used to perform quality assurance (QA) tests. These phantoms are not only expensive, but suitably trained personnel are needed to successfully use them and interpret the results. The question arose as to whether a single universal phantom could be designed and applied to all of the X‐ray imaging modalities. A universal phantom would reduce initial procurement cost, possibly reduce the time spent on QA procedures and simplify training of staff on the single device. The aim of the study was to design and manufacture a prototype of a universal phantom, suitable for image quality assurance in general X‐rays, fluoroscopy, mammography, and CT scanning. The universal phantom should be easy to use and would enable automatic data analysis, pass/fail reporting, and corrective action recommendation. In addition, a universal phantom would especially be of value in low‐income countries where finances and human resources are limited. The design process included a thorough investigation of commercially available phantoms. Image quality parameters necessary for image quality assurance in the different X‐ray imaging modalities were determined. Based on information obtained from the above‐mentioned investigations, a prototype of a universal phantom was developed, keeping ease of use and reduced cost in mind. A variety of possible phantom housing and insert materials were investigated, considering physical properties, machinability, and cost. A three‐dimensional computer model of the first phantom prototype was used to manufacture the prototype housing and inserts. Some of the inserts were 3D‐printed, others were machined from different materials. The different components were assembled to form the first prototype of the universal X‐ray imaging phantom. The resulting prototype of the universal phantom conformed to the aims of a single phantom for multiple imaging modalities, which would be easy to use and manufacture at a reduced cost. A PCT International Patent Application No. PCT/IB2016/051165 has been filed for this technology.PACS number(s): 87.57.C, 87.59.‐e

show abstract

“…An alternative is to use computational observers. [28][29][30][31] A widely used model in medical imaging is the channelized Hotelling observer (CHO). The CHO models the human observer using spatial frequency-selective filters, or channels, and optimal linear statistics to build a template for signal detection.…”

Section: Introductionmentioning

confidence: 99%

Computational and human observer image quality evaluation of low dose, knowledge‐based CT iterative reconstruction

et al. 2015

View full text Add to dashboard Cite

Purpose: Aims in this study are to (1) develop a computational model observer which reliably tracks the detectability of human observers in low dose computed tomography (CT) images reconstructed with knowledge-based iterative reconstruction (IMR™, Philips Healthcare) and filtered back projection (FBP) across a range of independent variables, (2) use the model to evaluate detectability trends across reconstructions and make predictions of human observer detectability, and (3) perform human observer studies based on model predictions to demonstrate applications of the model in CT imaging. Methods: Detectability (d ′ ) was evaluated in phantom studies across a range of conditions. Images were generated using a numerical CT simulator. Trained observers performed 4-alternative forced choice (4-AFC) experiments across dose (1.3, 2.7, 4.0 mGy), pin size (4, 6, 8 mm), contrast (0.3%, 0.5%, 1.0%), and reconstruction (FBP, IMR), at fixed display window. A five-channel Laguerre-Gauss channelized Hotelling observer (CHO) was developed with internal noise added to the decision variable and/or to channel outputs, creating six different internal noise models. Semianalytic internal noise computation was tested against Monte Carlo and used to accelerate internal noise parameter optimization. Model parameters were estimated from all experiments at once using maximum likelihood on the probability correct, P C . Akaike information criterion (AIC) was used to compare models of different orders. The best model was selected according to AIC and used to predict detectability in blended FBP-IMR images, analyze trends in IMR detectability improvements, and predict dose savings with IMR. Predicted dose savings were compared against 4-AFC study results using physical CT phantom images. Results: Detection in IMR was greater than FBP in all tested conditions. The CHO with internal noise proportional to channel output standard deviations, Model-k4, showed the best trade-off between fit and model complexity according to AIC c . With parameters fixed, the model reasonably predicted detectability of human observers in blended FBP-IMR images. Semianalytic internal noise computation gave results equivalent to Monte Carlo, greatly speeding parameter estimation. Using Model-k4, the authors found an average detectability improvement of 2.7 ± 0.4 times that of FBP. IMR showed greater improvements in detectability with larger signals and relatively consistent improvements across signal contrast and x-ray dose. In the phantom tested, Model-k4 predicted an 82% dose reduction compared to FBP, verified with physical CT scans at 80% reduced dose. Conclusions: IMR improves detectability over FBP and may enable significant dose reductions. A channelized Hotelling observer with internal noise proportional to channel output standard deviation agreed well with human observers across a wide range of variables, even across reconstructions with drastically different image characteristics. Utility of the model observer was demonstrated by predicting the effect of im...

show abstract

Automated assessment of low contrast sensitivity for CT systems using a model observer

Cited by 48 publications

References 24 publications

Comparison between human and model observer performance in low-contrast detection tasks in CT images: application to images reconstructed with filtered back projection and iterative algorithms

Comparison between human and model observer performance in low-contrast detection tasks in CT images: application to images reconstructed with filtered back projection and iterative algorithms

Development of a universal medical X‐ray imaging phantom prototype

Computational and human observer image quality evaluation of low dose, knowledge‐based CT iterative reconstruction

Contact Info

Product

Resources

About