A method to produce and validate a digitally reconstructed radiograph-based computer simulation for optimisation of chest radiographs acquired with a computed radiography imaging system

Virtual reality simulator's effectiveness on the spine procedure education for trainee: a randomized controlled trial

BACKGROUND

Since the onset of the coronavirus disease 2019 pandemic, virtual simulation has emerged as an alternative to traditional teaching methods as it can be employed within the recently established contact-minimizing guidelines. This prospective education study aimed to develop a virtual reality simulator for a lumbar transforaminal epidural block (LTFEB) and demonstrate its efficacy.

METHODS

We developed a virtual reality simulator using patient image data processing, virtual X-ray generation, spatial registration, and virtual reality technology. For a realistic virtual environment, a procedure room, surgical table, C-arm, and monitor were created. Using the virtual C-arm, the X-ray images of the patient's anatomy, the needle, and indicator were obtained in real-time. After the simulation, the trainees could receive feedback by adjusting the visibility of structures such as skin and bones. The training of LTFEB using the simulator was evaluated using 20 inexperienced trainees. The trainees' procedural time, rating score, number of C-arm taken, and overall satisfaction were recorded as primary outcomes.

RESULTS

The group using the simulator showed a higher global rating score (P = 0.014), reduced procedural time (P = 0.025), reduced number of C-arm uses (P = 0.001), and higher overall satisfaction score (P = 0.007).

CONCLUSIONS

We created an accessible and effective virtual reality simulator that can be used to teach inexperienced trainees LTFEB without radiation exposure. The results of this study indicate that the proposed simulator will prove to be a useful aid for teaching LTFEB.

X23D-Intraoperative 3D Lumbar Spine Shape Reconstruction Based on Sparse Multi-View X-ray Data

Visual assessment based on intraoperative 2D X-rays remains the predominant aid for intraoperative decision-making, surgical guidance, and error prevention. However, correctly assessing the 3D shape of complex anatomies, such as the spine, based on planar fluoroscopic images remains a challenge even for experienced surgeons. This work proposes a novel deep learning-based method to intraoperatively estimate the 3D shape of patients' lumbar vertebrae directly from sparse, multi-view X-ray data. High-quality and accurate 3D reconstructions were achieved with a learned multi-view stereo machine approach capable of incorporating the X-ray calibration parameters in the neural network. This strategy allowed a priori knowledge of the spinal shape to be acquired while preserving patient specificity and achieving a higher accuracy compared to the state of the art. Our method was trained and evaluated on 17,420 fluoroscopy images that were digitally reconstructed from the public CTSpine1K dataset. As evaluated by unseen data, we achieved an 88% average F1 score and a 71% surface score. Furthermore, by utilizing the calibration parameters of the input X-rays, our method outperformed a counterpart method in the state of the art by 22% in terms of surface score. This increase in accuracy opens new possibilities for surgical navigation and intraoperative decision-making solely based on intraoperative data, especially in surgical applications where the acquisition of 3D image data is not part of the standard clinical workflow.

A new 2D-3D registration gold-standard dataset for the hip joint based on uncertainty modeling

Purpose

Estimation of the accuracy of 2D‐3D registration is paramount for a correct evaluation of its outcome in both research and clinical studies. Publicly available datasets with standardized evaluation methodology are necessary for validation and comparison of 2D‐3D registration techniques. Given the large use of 2D‐3D registration in biomechanics, we introduced the first gold standard validation dataset for computed tomography (CT)‐to‐x‐ray registration of the hip joint, based on fluoroscopic images with large rotation angles. As the ground truth computed with fiducial markers is affected by localization errors in the image datasets, we proposed a new methodology based on uncertainty propagation to estimate the accuracy of a gold standard dataset.

Methods

The gold standard dataset included a 3D CT scan of a female hip phantom and 19 2D fluoroscopic images acquired at different views and voltages. The ground truth transformations were estimated based on the corresponding pairs of extracted 2D and 3D fiducial locations. These were assumed to be corrupted by Gaussian noise, without any restrictions of isotropy. We devised the multiple projective points criterion (MPPC) that jointly optimizes the transformations and the noisy 3D fiducial locations for all views. The accuracy of the transformations obtained with the MPPC was assessed in both synthetic and real experiments using different formulations of the target registration error (TRE), including a novel formulation of the TRE (uTRE) derived from the uncertainty analysis of the MPPC.

Results

The proposed MPPC method was statistically more accurate compared to the validation methods for 2D‐3D registration that did not optimize the 3D fiducial positions or wrongly assumed the isotropy of the noise. The reported results were comparable to previous published works of gold standard datasets. However, a formulation of the TRE commonly found in these gold standard datasets was found to significantly miscalculate the true TRE computed in synthetic experiments with known ground truths. In contrast, the uncertainty‐based uTRE was statistically closer to the true TRE.

Conclusions

We proposed a new gold standard dataset for the validation of CT‐to‐X‐ray registration of the hip joint. The gold standard transformations were derived from a novel method modeling the uncertainty in extracted 2D and 3D fiducials. Results showed that considering possible noise anisotropy and including corrupted 3D fiducials in the optimization resulted in improved accuracy of the gold standard. A new uncertainty‐based formulation of the TRE also appeared as a good alternative to the unknown true TRE that has been replaced in previous works by an alternative TRE not fully reflecting the gold standard accuracy.

Quantifying lung and diaphragm morphology using radiographs in normative pediatric subjects, and predicting CT-derived lung volume

OBJECTIVE

To quantify the effect of age on two-dimensional (2D) radiographic lung and diaphragm morphology and determine if 2D radiographic lung measurements can be used to estimate computer tomography (CT)-derived lung volume in normative pediatric subjects.

MATERIALS AND METHODS

Digitally reconstructed radiographs (DRRs) were created using retrospective chest CT scans from 77 pediatric male and female subjects aged birth to 19 years. 2D lung and diaphragm measurements were made on the DRRs using custom MATLAB code, and Spearman correlations and exponential regression equations were used to relate 2D measurements with age. In addition, 3D lung volumes were segmented using CT scans, and power regression equations were fitted to predict each lung's CT-derived volume from 2D lung measurements. The coefficient of determination (R² ) and standard error of the estimate (SEE) were used to assess the precision of the predictive equations with p < .05 indicating statistical significance.

RESULTS

All 2D radiographic lung and diaphragm measurements showed statistically significant positive correlations with age (p < .01), including lung major axis (Spearman rho ≥  0.90). Precise estimations of CT-derived lung volumes can be made using 2D lung measurements (R²  ≥ 0.95), including lung major axis (R²  ≥ 0.97).

INTERPRETATIONS

The reported pediatric age-specific reference data on 2D lung and diaphragm morphology and growth rates could be clinically used to identify lung and diaphragm pathologies during chest X-ray evaluations. The simple, precise, and clinically adaptable radiographic method for estimating CT-derived lung volumes may be used when pulmonary function tests are not readily available or difficult to perform.

RealDRR - Rendering of realistic digitally reconstructed radiographs using locally trained image-to-image translation

INTRODUCTION

Digitally reconstructed radiographs (DRRs) represent valuable patient-specific pre-treatment training data for tumor tracking algorithms. However, using current rendering methods, the similarity of the DRRs to real X-ray images is limited, requires time-consuming measurements and/or are computationally expensive. In this study we present RealDRR, a novel framework for highly realistic and computationally efficient DRR rendering.

MATERIALS AND METHODS

RealDRR consists of two components applied sequentially to render a DRR. First, a raytracer is applied for forward projection from 3D CT data to a 2D image. Second, a conditional Generative Adverserial Network (cGAN) is applied to translate the 2D forward projection to a realistic 2D DRR. The planning CT and CBCT projections from a CIRS thorax phantom and 6 radiotherapy patients (3 prostate, 3 brain) were split in training and test sets for evaluating the intra-patient, inter-patient and inter-anatomical region generalization performance of the trained framework. Several image similarity metrics, as well as a verification based on template matching, were used between the rendered DRRs and respective CBCT projections in the test sets, and results were compared to those of a current state-of-the-art DRR rendering method.

RESULTS

When trained on 800 CBCT projection images from two patients and tested on a third unseen patient from either anatomical region, RealDRR outperformed the current state-of-the-art with statistical significance on all metrics (two-sample t-test, p < 0.05). Once trained, the framework is able to render 100 highly realistic DRRs in under two minutes.

CONCLUSION

A novel framework for realistic and efficient DRR rendering was proposed. As the framework requires a reasonable amount of computational resources, the internal parameters can be tailored to imaging systems and protocols through on-site training on retrospective imaging data.

Optimisation of tube voltage range (kVp) for AP abdomen, pelvis and spine imaging of average patients with a digital radiography (DR) imaging system using a computer simulator

Objectives:

To investigate via computer simulation, an optimised tube voltage (kVp) range for caesium iodide (CsI)-based digital radiography (DR) of the abdomen, pelvis and lumbar spine.

Methods:

Software capable of simulating abdomen, pelvis and spine radiographs was used. Five evaluators graded clinical image criteria in images of 20 patients at tube voltages ranging from 60 to 120 kVp in 10 kVp increments. These criteria were scored blindly against the same patient reconstructed at a specific reference kVp. Linear mixed effects analysis was used to evaluate image scores for each criterion and test for statistical significance.

Results:

Score was dependent on tube voltage and image criteria; both were statistically significant. All criteria for all anatomies scored very poorly at 60 kVp. Scores for abdomen, pelvis and spine imaging peaked at 70, 70 and 100 kVp, respectively, but other kVp values were not significantly poorer.

Conclusions:

Results indicate optimum tube voltages of 70 kVp for abdomen and pelvis (with an optimum range 70–120 kVp), and 100 kVp (optimum range 80–120 kVp) for lumbar spine.

Advances in knowledge:

There are no recommendations for optimised tube voltage parameters for DR abdomen, pelvis or lumbar spine imaging. This study has investigated and recommended an optimal tube voltage range.

Volume Visualization for Improving CT Lung Nodule Detection

Inspired by the outstanding performance of deep convolutional neural networks (CNNs), nowadays modern computer-aided detection (CAD) systems for CT lung nodules generally delve into 2D or 3D CNNs directly without considering traditional image preprocessing techniques. However, detection of large pulmonary nodules and masses are computationally challenging, especially for 3D CNNs. In this paper, we examine the possibility of using volume visualized CT thin-slab images with 2D CNNs to reduce computation complexity and improve CAD performance. We tested 4 types of images: original 2D CT, 2D projection of thin slabs, mixture by arranging original and projection in different color channels, and mixture by the pixelwise maximum intensity of original CT and projection. We evaluated these images on a dataset of 30 CT scans with 30 different-sized nodules and masses on GoogLeNet via a transfer learning and cross validation paradigm. We found that projection visualization alone had a better or equal area-under curve score for all the different-sized nodules and masses. However, mixture by the maximum of CT and projection demonstrated a preferred performance with a true positive rate of 0.8 and a false positive rate of 0.046 in detecting large nodules and masses.

Assessment of Image Quality in Digital Radiographs Submitted for Hip Dysplasia Screening

Digital radiography is widely seen to be forgiving of poor exposure technique and to provide consistent high quality diagnostic images. Optimal quality images are however not universal; sub-optimal images are encountered. Evaluators on hip dysplasia schemes encounter images from multiple practices produced on equipment from multiple manufacturers. For images submitted to the Danish Kennel Club for hip dysplasia screening, a range of quality is seen and the evaluators are of the impression that variations in image quality area associated with particular equipment. This study was undertaken to test the hypothesis that there is an association between image quality in digital radiography and the manufacturer of the detector equipment, and to demonstrate the applicability of visual grading analysis (VGA) for image quality evaluation in veterinary practice. Data from 16,360 digital images submitted to the Danish Kennel Club were used to generate the hypothesis that there is an association between detector manufacturer and image quality and to create groups for VGA. Image quality in a subset of 90 images randomly chosen from 6 manufacturers to represent high and low quality images, was characterized using VGA and the results used to test for an association between image quality and system manufacturer. The range of possible scores in the VGA was −2 to +2 (higher scores are better). The range of the VGA scores for the images in the low image quality group (n = 45) was −1.73 to +0.67, (median −1.2). Images in the high image quality group (n = 44) ranged from −1.52 to +0.53, (median −0.53). This difference was statistically significant (p < 0.001). The study shows an association between VGA scores of image quality and detector manufacturer. Possible causes may be that imaging hardware and/or software are not equal in terms of quality, that the level of support sought and given differs between systems, or a combination of the two. Clinicians purchasing equipment should be mindful that image quality can differ across systems. VGA is practical for veterinarians to compare image quality between systems or within a system over time.

Use of a computer simulator to investigate optimized tube voltage for chest imaging of average patients with a digital radiography (DR) imaging system

Objective:

The aim of this study was to investigate via computer simulation a proposed improvement to clinical practice by deriving an optimized tube voltage (kVp) range for digital radiography (DR) chest imaging.

Methods:

A digitally reconstructed radiograph algorithm was used which was capable of simulating DR chest radiographs containing clinically relevant anatomy. Five experienced image evaluators graded clinical image criteria, i.e. overall quality, rib, lung, hilar, spine, diaphragm and lung nodule in images of 20 patients at tube voltages across the diagnostic energy range. These criteria were scored against corresponding images of the same patient reconstructed at a specific reference kVp. Evaluators were blinded to kVp. Evaluator score for each criterion was modelled with a linear mixed effects algorithm and compared with the score for the reference image.

Results:

Score was dependent on tube voltage and image criteria in a statistically significant manner for both. Overall quality, hilar, diaphragm and spine criteria performed poorly at low and high tube voltages, peaking at 80–100 kVp. Lung and lung nodule demonstrated little variation. Rib demonstrated superiority at low kVp.

Conclusion:

A virtual clinical trial has been performed with simulated chest DR images. Results indicate mid-range tube voltages of 80–100 kVp are optimum for average adults.

Advances in knowledge:

There are currently no specific recommendations for optimized tube voltage parameters for DR chest imaging. This study, validated with images containing realistic anatomical noise, has investigated and recommended an optimal tube voltage range.

An investigation into the validity of utilising the CDRAD 2.0 phantom for optimisation studies in digital radiography

OBJECTIVES

To determine if a relationship exists between low contrast detail (LCD) detectability using the CDRAD 2.0 phantom, visual measures of image quality (IQ) and simulated lesion visibility (LV) when performing digital chest radiography (CXR).

METHODS

Using a range of acquisition parameters, a CDRAD 2.0 phantom was used to acquire a set of images with different levels of image quality. LCD detectability using the CDRAD 2.0 phantom, represented by an image quality figure inverse (IQF_inv) metric, was determined using the phantom analyser software. A Lungman chest phantom was loaded with two simulated lesions, of different sizes/placed in different locations, and was imaged using the same acquisition factors as the CDRAD 2.0 phantom. A relative visual grading analysis (VGA) was used by seven observers for IQ and LV evaluation of the Lungman images. Correlations between IQF_inv, IQ and LV were investigated.

RESULTS

Pearson's correlation demonstrated a strong positive correlation (r = 0.91; p < 0.001) between the IQ and the IQF_inv. Spearman's correlation showed a good positive correlation (r = 0.79; p < 0.001) and (r = 0.68; p < 0.001) between the IQF_inv and the LV for the first lesion (left upper lobe) and the second lesion (right middle lobe), respectively.

CONCLUSIONS

From results presented in this study, the automated evaluation of LCD detectability using CDRAD 2.0 phantom is likely to be a suitable option for IQ and LV evaluation in digital CXR optimisation studies. Advances in knowledge: This research establishes the potential of the CDRAD 2.0 phantom in digital CXR optimisation studies.

A method to incorporate the effect of beam quality on image noise in a digitally reconstructed radiograph (DRR) based computer simulation for optimisation of digital radiography

The use of computer simulated digital x-radiographs for optimisation purposes has become widespread in recent years. To make these optimisation investigations effective, it is vital simulated radiographs contain accurate anatomical and system noise. Computer algorithms that simulate radiographs based solely on the incident detector x-ray intensity ('dose') have been reported extensively in the literature. However, while it has been established for digital mammography that x-ray beam quality is an important factor when modelling noise in simulated images there are no such studies for diagnostic imaging of the chest, abdomen and pelvis. This study investigates the influence of beam quality on image noise in a digital radiography (DR) imaging system, and incorporates these effects into a digitally reconstructed radiograph (DRR) computer simulator. Image noise was measured on a real DR imaging system as a function of dose (absorbed energy) over a range of clinically relevant beam qualities. Simulated 'absorbed energy' and 'beam quality' DRRs were then created for each patient and tube voltage under investigation. Simulated noise images, corrected for dose and beam quality, were subsequently produced from the absorbed energy and beam quality DRRs, using the measured noise, absorbed energy and beam quality relationships. The noise images were superimposed onto the noiseless absorbed energy DRRs to create the final images. Signal-to-noise measurements in simulated chest, abdomen and spine images were within 10% of the corresponding measurements in real images. This compares favourably to our previous algorithm where images corrected for dose only were all within 20%.

A GPU Simulation Tool for Training and Optimisation in 2D Digital X-Ray Imaging

Conventional radiology is performed by means of digital detectors, with various types of technology and different performance in terms of efficiency and image quality. Following the arrival of a new digital detector in a radiology department, all the staff involved should adapt the procedure parameters to the properties of the detector, in order to achieve an optimal result in terms of correct diagnostic information and minimum radiation risks for the patient. The aim of this study was to develop and validate a software capable of simulating a digital X-ray imaging system, using graphics processing unit computing. All radiological image components were implemented in this application: an X-ray tube with primary beam, a virtual patient, noise, scatter radiation, a grid and a digital detector. Three different digital detectors (two digital radiography and a computed radiography systems) were implemented. In order to validate the software, we carried out a quantitative comparison of geometrical and anthropomorphic phantom simulated images with those acquired. In terms of average pixel values, the maximum differences were below 15%, while the noise values were in agreement with a maximum difference of 20%. The relative trends of contrast to noise ratio versus beam energy and intensity were well simulated. Total calculation times were below 3 seconds for clinical images with pixel size of actual dimensions less than 0.2 mm. The application proved to be efficient and realistic. Short calculation times and the accuracy of the results obtained make this software a useful tool for training operators and dose optimisation studies.

Investigating the use of an antiscatter grid in chest radiography for average adults with a computed radiography imaging system

OBJECTIVE

The aim of this study was to investigate via simulation a proposed change to clinical practice for chest radiography. The validity of using a scatter rejection grid across the diagnostic energy range (60-125 kVp), in conjunction with appropriate tube current-time product (mAs) for imaging with a computed radiography (CR) system was investigated.

METHODS

A digitally reconstructed radiograph algorithm was used, which was capable of simulating CR chest radiographs with various tube voltages, receptor doses and scatter rejection methods. Four experienced image evaluators graded images with a grid (n = 80) at tube voltages across the diagnostic energy range and varying detector air kermas. These were scored against corresponding images reconstructed without a grid, as per current clinical protocol.

RESULTS

For all patients, diagnostic image quality improved with the use of a grid, without the need to increase tube mAs (and therefore patient dose), irrespective of the tube voltage used. Increasing tube mAs by an amount determined by the Bucky factor made little difference to image quality.

CONCLUSION

A virtual clinical trial has been performed with simulated chest CR images. RESULTS indicate that the use of a grid improves diagnostic image quality for average adults, without the need to increase tube mAs, even at low tube voltages.

ADVANCES IN KNOWLEDGE

Validated with images containing realistic anatomical noise, it is possible to improve image quality by utilizing grids for chest radiography with CR systems without increasing patient exposure. Increasing tube mAs by an amount determined by the Bucky factor is not justified.

Accelerating DRR generation using Fourier slice theorem on the GPU

Digitally Reconstructed Radiographs (DRRs) play a vital role in medical imaging procedures and radiotherapy applications. They allow the continuous monitoring of patient positioning during image guided therapies using multi-dimensional image registration. Conventional generation of DRRs using spatial domain algorithms such as ray casting is associated with computational complexity of O(N(3)). Fourier slice theorem is an alternative approach for generating the DRRs in the k-space with reduced time complexity. In this work, we present a high performance, scalable, and optimized DRR generation pipeline on the Graphics Processing Unit (GPU). The strong scaling performance of the presented pipeline is investigated and demonstrated using two contemporary GPUs. Our pipeline is capable of generating DRRs for 512(3) volumes in less than a milli-second.

A digitally reconstructed radiograph algorithm calculated from first principles

PURPOSE

To develop an algorithm for computing realistic digitally reconstructed radiographs (DRRs) that match real cone-beam CT (CBCT) projections with no artificial adjustments.

METHODS

The authors used measured attenuation data from cone-beam CT projection radiographs of different materials to obtain a function to convert CT number to linear attenuation coefficient (LAC). The effects of scatter, beam hardening, and veiling glare were first removed from the attenuation data. Using this conversion function the authors calculated the line integral of LAC through a CT along rays connecting the radiation source and detector pixels with a ray-tracing algorithm, producing raw DRRs. The effects of scatter, beam hardening, and veiling glare were then included in the DRRs through postprocessing.

RESULTS

The authors compared actual CBCT projections to DRRs produced with all corrections (scatter, beam hardening, and veiling glare) and to uncorrected DRRs. Algorithm accuracy was assessed through visual comparison of projections and DRRs, pixel intensity comparisons, intensity histogram comparisons, and correlation plots of DRR-to-projection pixel intensities. In general, the fully corrected algorithm provided a small but nontrivial improvement in accuracy over the uncorrected algorithm. The authors also investigated both measurement- and computation-based methods for determining the beam hardening correction, and found the computation-based method to be superior, as it accounted for nonuniform bowtie filter thickness. The authors benchmarked the algorithm for speed and found that it produced DRRs in about 0.35 s for full detector and CT resolution at a ray step-size of 0.5 mm.

CONCLUSIONS

The authors have demonstrated a DRR algorithm calculated from first principles that accounts for scatter, beam hardening, and veiling glare in order to produce accurate DRRs. The algorithm is computationally efficient, making it a good candidate for iterative CT reconstruction techniques that require a data fidelity term based on the matching of DRRs and projections.

Correlation of the clinical and physical image quality in chest radiography for average adults with a computed radiography imaging system

OBJECTIVE

The purpose of this study was to examine the correlation between the quality of visually graded patient (clinical) chest images and a quantitative assessment of chest phantom (physical) images acquired with a computed radiography (CR) imaging system.

METHODS

The results of a previously published study, in which four experienced image evaluators graded computer-simulated postero-anterior chest images using a visual grading analysis scoring (VGAS) scheme, were used for the clinical image quality measurement. Contrast-to-noise ratio (CNR) and effective dose efficiency (eDE) were used as physical image quality metrics measured in a uniform chest phantom. Although optimal values of these physical metrics for chest radiography were not derived in this work, their correlation with VGAS in images acquired without an antiscatter grid across the diagnostic range of X-ray tube voltages was determined using Pearson's correlation coefficient.

RESULTS

Clinical and physical image quality metrics increased with decreasing tube voltage. Statistically significant correlations between VGAS and CNR (R=0.87, p<0.033) and eDE (R=0.77, p<0.008) were observed.

CONCLUSION

Medical physics experts may use the physical image quality metrics described here in quality assurance programmes and optimisation studies with a degree of confidence that they reflect the clinical image quality in chest CR images acquired without an antiscatter grid.

ADVANCES IN KNOWLEDGE

A statistically significant correlation has been found between the clinical and physical image quality in CR chest imaging. The results support the value of using CNR and eDE in the evaluation of quality in clinical thorax radiography.

Use of a digitally reconstructed radiograph-based computer simulation for the optimisation of chest radiographic techniques for computed radiography imaging systems

OBJECTIVES

The purpose of this study was to derive an optimum radiographic technique for computed radiography (CR) chest imaging using a digitally reconstructed radiograph computer simulator. The simulator is capable of producing CR chest radiographs of adults with various tube potentials, receptor doses and scatter rejection.

METHODS

Four experienced image evaluators graded images of average and obese adult patients at different potentials (average-sized, n=50; obese, n=20), receptor doses (n=10) and scatter rejection techniques (average-sized, n=20; obese, n=20). The quality of the images was evaluated using visually graded analysis. The influence of rib contrast was also assessed.

RESULTS

For average-sized patients, image quality improved when tube potential was reduced compared with the reference (102 kVp). No scatter rejection was indicated. For obese patients, it has been shown that an antiscatter grid is indicated, and should be used in conjunction with as low a tube potential as possible (while allowing exposure times <20 ms). It is also possible to reduce receptor air kerma by 50% without adversely influencing image quality. Rib contrast did not interfere at any tube potential.

CONCLUSIONS

A virtual clinical trial has been performed with simulated chest CR images. Results indicate that low tube potentials (<102 kVp) are optimal for average and obese adults, the former acquired without scatter rejection, the latter with an anti-scatter grid. Lower receptor (and therefore patient doses) than those used clinically are possible while maintaining adequate image quality.