Quantitative accuracy of CT numbers: Theoretical analyses and experimental studies

A Review on the Use of Imaging Biomarkers in Oncology Clinical Trials: Quality Assurance Strategies for Technical Validation

Imaging biomarkers (IBs) have been proposed in medical literature that exploit images in a quantitative way, going beyond the visual assessment by an imaging physician. These IBs can be used in the diagnosis, prognosis, and response assessment of several pathologies and are very often used for patient management pathways. In this respect, IBs to be used in clinical practice and clinical trials have a requirement to be precise, accurate, and reproducible. Due to limitations in imaging technology, an error can be associated with their value when considering the entire imaging chain, from data acquisition to data reconstruction and subsequent analysis. From this point of view, the use of IBs in clinical trials requires a broadening of the concept of quality assurance and this can be a challenge for the responsible medical physics experts (MPEs). Within this manuscript, we describe the concept of an IB, examine some examples of IBs currently employed in clinical practice/clinical trials and analyze the procedure that should be carried out to achieve better accuracy and reproducibility in their use. We anticipate that this narrative review, written by the components of the EFOMP working group on "the role of the MPEs in clinical trials"-imaging sub-group, can represent a valid reference material for MPEs approaching the subject.

Correcting statistical CT number biases without access to raw detector counts: Applications to high spatial resolution photon counting CT imaging

BACKGROUND

Due to the nonlinear nature of the logarithmic operation and the stochastic nature of photon counts (N), sinogram data of photon counting detector CT (PCD-CT) are intrinsically biased, which leads to statistical CT number biases. When raw counts are available, nearly unbiased statistical estimators for projection data were developed recently to address the CT number bias issue. However, for most clinical PCD-CT systems, users' access to raw detector counts is limited. Therefore, it remains a challenge for end users to address the CT number bias issue in clinical applications.

PURPOSE

To develop methods to correct statistical biases in PCD-CT without requiring access to raw PCD counts.

METHODS

(1) The sample variance of air-only post-log sinograms was used to estimate air-only detector counts,N ¯ 0 $\bar{N}_0$ . (2) If the post-log sinogram data, y, is available, then N of each detector pixel was estimated usingN = N ¯ 0 e - y $N = \bar{N}_0 \, \mathrm{e}^{-y}$ . Once N was estimated, a closed-form analytical bias correction was applied to the sinogram. (3) If a patient's post-log sinogram data are not archived, a forward projection of the bias-contaminated CT image was used to perform a first-order bias correction. Both the proposed sinogram domain- and image domain-based bias correction methods were validated using experimental PCD-CT data.

RESULTS

Experimental results demonstrated that both sinogram domain- and image domain-based bias correction methods enabled reduced-dose PCD-CT images to match the CT numbers of reference-standard images within [-5, 5] HU. In contrast, uncorrected reduced-dose PCD-CT images demonstrated biases ranging from -25 to 55 HU, depending on the material. No increase in image noise or spatial resolution degradation was observed using the proposed methods.

CONCLUSIONS

CT number bias issues can be effectively addressed using the proposed sinogram or image domain method in PCD-CT, allowing PCD-CT acquired at different radiation dose levels to have consistent CT numbers desired for quantitative imaging.

Impact of Noise Level on the Accuracy of Automated Measurement of CT Number Linearity on ACR CT and Computational Phantoms

Background

Methods for segmentation, i.e., Full-segmentation (FS) and Segmentation-rotation (SR), are proposed for maintaining Computed Tomography (CT) number linearity. However, their effectiveness has not yet been tested against noise.

Objective

This study aimed to evaluate the influence of noise on the accuracy of CT number linearity of the FS and SR methods on American College of Radiology (ACR) CT and computational phantoms.

Material and Methods

This experimental study utilized two phantoms, ACR CT and computational phantoms. An ACR CT phantom was scanned by a 128-slice CT scanner with various tube currents from 80 to 200 mA to acquire various noises, with other constant parameters. The computational phantom was added by different Gaussian noises between 20 and 120 Hounsfield Units (HU). The CT number linearity was measured by the FS and SR methods, and the accuracy of CT number linearity was computed on two phantoms.

Results

The two methods successfully segmented both phantoms at low noise, i.e., less than 60 HU. However, segmentation and measurement of CT number linearity are not accurate on a computational phantom using the FS method for more than 60-HU noise. The SR method is still accurate up to 120 HU of noise.

Conclusion

The SR method outperformed the FS method to measure the CT number linearity due to its endurance in extreme noise.

Liver Attenuation Assessment in Reduced Radiation Chest Computed Tomography

OBJECTIVE

This study aimed to evaluate the reliability of liver and spleen Hounsfield units (HU) measurements in reduced radiation computed tomography (RRCT) of the chest within the sub-millisievert range.

METHODS

We performed a prospective, institutional review board-approved study of accrued patients who underwent unenhanced normal-dose chest CT (NDCT) and with an average radiation dose of less than 5% of NDCT. In-house artificial intelligence-based denoising methods produced 2 denoised RRCT (dRRCT) series. Hepatic and splenic attenuations were measured on all 4 series: NDCT, RRCT, dRRCT1, and dRRCT2. Statistical analyses assessed the differences between the HU measurements of the liver and spleen in RRCTs and NDCT. As a test case, we assessed the performance of RRCTs for fatty liver detection, considering NDCT to be the reference standard.

RESULTS

Wilcoxon test compared liver and spleen attenuation in the 72 patients included in our cohort. The liver attenuation in NDCT (median, 59.38 HU; interquartile range, 55.00-66.06 HU) was significantly different from the attenuation in RRCT, dRRCT1, and dRRCT2 (median, 63.63, 42.00, and 33.67 HU; interquartile range, 56.19-67.19, 37.33-45.83, and 30.33-38.50 HU, respectively), all with a P value <0.01. Six patients (8.3%) were considered to have fatty liver on NDCT. The specificity, sensitivity, and accuracy of fatty liver detection by RRCT were greater than 98.5%, 50%, and 94.3%, respectively.

CONCLUSIONS

Attenuation measurements were significantly different between NDCT and RRCTs, but may still have diagnostic value in appreciating hepatosteastosis. Abdominal organ attenuation on RRCT protocols may differ from attenuation on NDCT and should be validated when new low-dose protocols are used.

The impacts of vertical off-centring, tube voltage, and phantom size on computed tomography numbers: An experimental study

INTRODUCTION

This experimental study explored the effect of vertical off-centring on computed tomography (CT) numbers in combination with various tube voltages and phantom sizes for two CT units.

METHODS

CIRS Model 062 Electron Density and system performance phantoms were imaged on Siemens Emotion 16-slice CT and GEMINI-GXL scanners, respectively. Uniformity and accuracy were evaluated as a function of vertical off-centring (20, 40, 60, and 80 mm above the gantry isocentre) using different water phantom sizes (18, 20, and 30 cm) and tube voltages (80, 90, 110, 120, 130 and 140 kVp).

RESULTS

Vertical off-centring and phantom size accounted for 92% of the recorded variance and the resultant change in CT numbers. The uniformity test recorded maximum changes of 14 and 27.2 HU for peripheral ROIs across the X- and Y-axes for an 80 mm phantom shift above the gantry isocentre on the GEMINI GXL and Siemens scanners, respectively. The absolute CT number differences between the superior and inferior ROIs were 13.7 HU for the 30 cm phantom and 4.8 HU for the 20 cm phantom for 80 mm vertical off-centring. The largest differences were observed at lower tube voltages.

CONCLUSIONS

It is essential to highlight the significance of CT number variation in clinical decision-making. Phantom off-centring affected the uniformity of these numbers, which were further impacted by the ROI position in this experimental study. CT number variation was more evident in peripheral phantom areas, lower tube voltages and larger phantom sizes.

IMPLICATIONS FOR PRACTICE

CT number is observed to be a variable under certain common conditions. This significantly impacts several applications where clinical decisions depend on CT number accuracy for tissue lesion characterisation.

Reconstruction of three-dimensional tomographic patient models for radiation dose modulation in CT from two scout views using deep learning

BACKGROUND

A tomographic patient model is essential for radiation dose modulation in x-ray computed tomography (CT). Currently, two-view scout images (also known as topograms) are used to estimate patient models with relatively uniform attenuation coefficients. These patient models do not account for the detailed anatomical variations of human subjects, and thus, may limit the accuracy of intraview or organ-specific dose modulations in emerging CT technologies.

PURPOSE

The purpose of this work was to show that 3D tomographic patient models can be generated from two-view scout images using deep learning strategies, and the reconstructed 3D patient models indeed enable accurate prescriptions of fluence-field modulated or organ-specific dose delivery in the subsequent CT scans.

METHODS

CT images and the corresponding two-view scout images were retrospectively collected from 4214 individual CT exams. The collected data were curated for the training of a deep neural network architecture termed ScoutCT-NET to generate 3D tomographic attenuation models from two-view scout images. The trained network was validated using a cohort of 55 136 images from 212 individual patients. To evaluate the accuracy of the reconstructed 3D patient models, radiation delivery plans were generated using ScoutCT-NET 3D patient models and compared with plans prescribed based on true CT images (gold standard) for both fluence-field-modulated CT and organ-specific CT. Radiation dose distributions were estimated using Monte Carlo simulations and were quantitatively evaluated using the Gamma analysis method. Modulated dose profiles were compared against state-of-the-art tube current modulation schemes. Impacts of ScoutCT-NET patient model-based dose modulation schemes on universal-purpose CT acquisitions and organ-specific acquisitions were also compared in terms of overall image appearance, noise magnitude, and noise uniformity.

RESULTS

The results demonstrate that (1) The end-to-end trained ScoutCT-NET can be used to generate 3D patient attenuation models and demonstrate empirical generalizability. (2) The 3D patient models can be used to accurately estimate the spatial distribution of radiation dose delivered by standard helical CTs prior to the actual CT acquisition; compared to the gold-standard dose distribution, 95.0% of the voxels in the ScoutCT-NET based dose maps have acceptable gamma values for 5 mm distance-to-agreement and 10% dose difference. (3) The 3D patient models also enabled accurate prescription of fluence-field modulated CT to generate a more uniform noise distribution across the patient body compared to tube current-modulated CT. (4) ScoutCT-NET 3D patient models enabled accurate prescription of organ-specific CT to boost image quality for a given body region-of-interest under a given radiation dose constraint.

CONCLUSION

3D tomographic attenuation models generated by ScoutCT-NET from two-view scout images can be used to prescribe fluence-field-modulated or organ-specific CT scans with high accuracy for the overall objective of radiation dose reduction or image quality improvement for a given imaging task.

The reliability of CT numbers as absolute values for diagnostic scanning, dental imaging, and radiation therapy simulation: A narrative review

BACKGROUND AND PURPOSE

The purpose of this review was to examine the reported factors that affect the reliability of Computed Tomography (CT) numbers and their impact on clinical applications in diagnostic scanning, dental imaging, and radiation therapy dose calculation.

METHODS

A comprehensive search of the literature was conducted using Medline (PubMed), Google Scholar, and Ovid databases which were searched using the keywords CT number variability, CT number accuracy and uniformity, tube voltage, patient positioning, patient off-centring, and size dependence. A narrative summary was used to compile the findings under the overarching theme.

DISCUSSION

A total of 47 articles were identified to address the aim of this review. There is clear evidence that CT numbers are highly dependent on the energy level applied based on the effective atomic number of the scanned tissue. Furthermore, body size and anatomical location have also indicated an influence on measured CT numbers, especially for high-density materials such as bone tissue and dental implants. Patient off-centring was reported during CT imaging, affecting dose and CT number reliability, which was demonstrated to be dependent on the shaping filter size.

CONCLUSION

CT number accuracy for all energy levels, body sizes, anatomical locations, and degrees of patient off-centring is observed to be a variable under certain common conditions. This has significant implications for several clinical applications. It is crucial for those involved in CT imaging to understand the limitations of their CT system to ensure radiologists and operators avoid potential pitfalls associated with using CT numbers as absolute values for diagnostic scanning, dental imaging, and radiation therapy dose calculation.

Photon-counting x-ray detectors for CT

The introduction of photon-counting detectors is expected to be the next major breakthrough in clinical x-ray computed tomography (CT). During the last decade, there has been considerable research activity in the field of photon-counting CT, in terms of both hardware development and theoretical understanding of the factors affecting image quality. In this article, we review the recent progress in this field with the intent of highlighting the relationship between detector design considerations and the resulting image quality. We discuss detector design choices such as converter material, pixel size, and readout electronics design, and then elucidate their impact on detector performance in terms of dose efficiency, spatial resolution, and energy resolution. Furthermore, we give an overview of data processing, reconstruction methods and metrics of imaging performance; outline clinical applications; and discuss potential future developments.

Effects of Patient Size and Radiation Dose on Iodine Quantification in Dual-Source Dual-Energy CT

RATIONALE AND OBJECTIVES

The purpose of this study was to investigate the potential effects of patient size and radiation dose on the accuracy of iodine quantification using dual-source dual-energy computed tomography (CT).

MATERIALS AND METHODS

Three phantoms representing different patient sizes were constructed, containing iodine inserts with concentrations from 0 to 20 mg/ml. Dual-energy CT scans were performed at six dose levels from 2 to 30 mGy. Iodine concentrations were measured using a three-material-decomposition algorithm and their accuracy was assessed.

RESULTS

In a small phantom, iodine quantification was accurate and consistent at all dose levels. In a medium phantom, minor underestimations were observed, and the results were consistent except at low dose. In the large phantom, more significant underestimation of iodine concentration was observed at higher doses (≥15 mGy), which was attributed to the beam-hardening effect. At lower doses, increasing upward bias was observed in the CT number, leading to significant overestimations of both iodine concentration and fat fraction, which was attributed to the photon-starvation effect. The severity of the latter effect was determined by mA instead of mAs, suggesting that the electronic noise, rather than the quantum noise, was responsible for the bias. Using higher kVp for the low-energy tube was found to alleviate these effects.

CONCLUSION

Reliable iodine quantification can be achieved using dual-source CT, but the result can be affected by patient size and dose rate. In large patients, biases may occur due to the beam-hardening and the photon-starvation effects, in which case higher dose rate and higher kVp are recommended to minimize these effects.

Deep-learning-based direct inversion for material decomposition

PURPOSE

To develop a convolutional neural network (CNN) that can directly estimate material density distribution from multi-energy computed tomography (CT) images without performing conventional material decomposition.

METHODS

The proposed CNN (denoted as Incept-net) followed the general framework of encoder-decoder network, with an assumption that local image information was sufficient for modeling the nonlinear physical process of multi-energy CT. Incept-net was implemented with a customized loss function, including an in-house-designed image-gradient-correlation (IGC) regularizer to improve edge preservation. The network consisted of two types of customized multibranch modules exploiting multiscale feature representation to improve the robustness over local image noise and artifacts. Inserts with various densities of different materials [hydroxyapatite (HA), iodine, a blood-iodine mixture, and fat] were scanned using a research photon-counting detector (PCD) CT with two energy thresholds and multiple radiation dose levels. The network was trained using phantom image patches only, and tested with different-configurations of full field-of-view phantom and in vivo porcine images. Furthermore, the nominal mass densities of insert materials were used as the labels in CNN training, which potentially provided an implicit mass conservation constraint. The Incept-net performance was evaluated in terms of image noise, detail preservation, and quantitative accuracy. Its performance was also compared to common material decomposition algorithms including least-square-based material decomposition (LS-MD), total-variation regularized material decomposition (TV-MD), and U-net-based method.

RESULTS

Incept-net improved accuracy of the predicted mass density of basis materials compared with the U-net, TV-MD, and LS-MD: the mean absolute error (MAE) of iodine was 0.66, 1.0, 1.33, and 1.57 mgI/cc for Incept-net, U-net, TV-MD, and LS-MD, respectively, across all iodine-present inserts (2.0-24.0 mgI/cc). With the LS-MD as the baseline, Incept-net and U-net achieved comparable noise reduction (both around 95%), both higher than TV-MD (85%). The proposed IGC regularizer effectively helped both Incept-net and U-net to reduce image artifact. Incept-net closely conserved the total mass densities (i.e., mass conservation constraint) in porcine images, which heuristically validated the quantitative accuracy of its outputs in anatomical background. In general, Incept-net performance was less dependent on radiation dose levels than the two conventional methods; with approximately 40% less parameters, the Incept-net achieved relatively improved performance than the comparator U-net, indicating that performance gain by Incept-net was not achieved by simply increasing network learning capacity.

CONCLUSION

Incept-net demonstrated superior qualitative image appearance, quantitative accuracy, and lower noise than the conventional methods and less sensitive to dose change. Incept-net generalized and performed well with unseen image structures and different material mass densities. This study provided preliminary evidence that the proposed CNN may be used to improve the material decomposition quality in multi-energy CT.