Dataset Information

Automatic Synthesis of Anthropomorphic Pulmonary CT Phantoms.

ABSTRACT: The great density and structural complexity of pulmonary vessels and airways impose limitations on the generation of accurate reference standards, which are critical in training and in the validation of image processing methods for features such as pulmonary vessel segmentation or artery-vein (AV) separations. The design of synthetic computed tomography (CT) images of the lung could overcome these difficulties by providing a database of pseudorealistic cases in a constrained and controlled scenario where each part of the image is differentiated unequivocally. This work demonstrates a complete framework to generate computational anthropomorphic CT phantoms of the human lung automatically. Starting from biological and image-based knowledge about the topology and relationships between structures, the system is able to generate synthetic pulmonary arteries, veins, and airways using iterative growth methods that can be merged into a final simulated lung with realistic features. A dataset of 24 labeled anthropomorphic pulmonary CT phantoms were synthesized with the proposed system. Visual examination and quantitative measurements of intensity distributions, dispersion of structures and relationships between pulmonary air and blood flow systems show good correspondence between real and synthetic lungs (p > 0.05 with low Cohen's d effect size and AUC values), supporting the potentiality of the tool and the usefulness of the generated phantoms in the biomedical image processing field.

SUBMITTER: Jimenez-Carretero D

PROVIDER: S-EPMC4711718 | biostudies-literature | 2016

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Automatic Synthesis of Anthropomorphic Pulmonary CT Phantoms.

Jimenez-Carretero Daniel D San Jose Estepar Raul R Diaz Cacio Mario M Ledesma-Carbayo Maria J MJ

PloS one 20160105 1

The great density and structural complexity of pulmonary vessels and airways impose limitations on the generation of accurate reference standards, which are critical in training and in the validation of image processing methods for features such as pulmonary vessel segmentation or artery-vein (AV) separations. The design of synthetic computed tomography (CT) images of the lung could overcome these difficulties by providing a database of pseudorealistic cases in a constrained and controlled scena ...[more]

PMID: 26731653

Similar Datasets

Project description:To present the design and procedure of an experimental method for acquiring densely sampled organ dose map for CT applications, based on optically stimulated luminescence (OSL) dosimeters "nanoDots" and standard ATOM anthropomorphic phantoms; and to provide the results of applying the method--a dose data set with good statistics for the comparison with Monte Carlo simulation result in the future.A standard ATOM phantom has densely located holes (in 3×3 cm or 1.5×1.5 cm grids), which are too small (5 mm in diameter) to host many types of dosimeters, including the nanoDots. The authors modified the conventional way in which nanoDots are used, by removing the OSL disks from the holders before inserting them inside a standard ATOM phantom for dose measurements. The authors solved three technical difficulties introduced by this modification: (1) energy dependent dose calibration for raw OSL readings; (2) influence of the brief background exposure of OSL disks to dimmed room light; (3) correct pairing between the dose readings and measurement locations. The authors acquired 100 dose measurements at various positions in the phantom, which was scanned using a clinical chest protocol with both angular and z-axis tube current modulations.Dose calibration was performed according to the beam qualities inside the phantom as determined from an established Monte Carlo model of the scanner. The influence of the brief exposure to dimmed room light was evaluated and deemed negligible. Pairing between the OSL readings and measurement locations was ensured by the experimental design. The organ doses measured for a routine adult chest scan protocol ranged from 9.4 to 18.8 mGy, depending on the composition, location, and surrounding anatomy of the organs. The dose distribution across different slices of the phantom strongly depended on the z-axis mA modulation. In the same slice, doses to the soft tissues other than the spinal cord demonstrated relatively small variations, with the maximum COV around 11.4%. This might be attributed to the angular mA modulation, the placement of the dosimeters, the chest cavity of the scanned region, and the size of the phantom. Doses to the spinal cord were consistently lower than those to other soft tissues.The method is suited for acquiring densely sampled organ dose maps, and can be used for studying dose distributions relevant to subject size, organ location, and clinical CT protocols.

Project description:The aims of this study were to characterize reinforced metal-oxide-semiconductor field-effect transistor (MOSFET) dosimeters to assess the measurement uncertainty, single exposure low-dose limit with acceptable accuracy, and the number of exposures required to attain the corresponding limit of the thermoluminescent dosimeters (TLD). The second aim was to characterize MOSFET dosimeter sensitivities for two dental photon energy ranges, dose dependency, dose rate dependency, and accumulated dose dependency. A further aim was to compare the performance of MOSFETs with those of TLDs in an anthropomorphic phantom head using a dentomaxillofacial CBCT device. The uncertainty was assessed by exposing 20 MOSFETs and a Barracuda MPD reference dosimeter. The MOSFET dosimeter sensitivities were evaluated for two photon energy ranges (50-90 kVp) using a constant dose and polymethylmethacrylate backscatter material. MOSFET and TLD comparative point-dose measurements were performed on an anthropomorphic phantom that was exposed with a clinical CBCT protocol. The MOSFET single exposure low dose limit (25% uncertainty, k = 2) was 1.69 mGy. An averaging of eight MOSFET exposures was required to attain the corresponding TLD (0.3 mGy) low-dose limit. The sensitivity was 3.09 ± 0.13 mV/mGy independently of the photon energy used. The MOSFET dosimeters did not present dose or dose rate sensitivity but, however, presented a 1% decrease of sensitivity per 1000 mV for accumulated threshold voltages between 8300 mV and 17500 mV. The point doses in an anthropomorphic phantom ranged for MOSFETs between 0.24 mGy and 2.29 mGy and for TLDs between 0.25 and 2.09 mGy, respectively. The mean difference was -8%. The MOSFET dosimeters presented statistically insignificant energy dependency. By averaging multiple exposures, the MOSFET dosimeters can achieve a TLD-comparable low-dose limit and constitute a feasible method for diagnostic dosimetry using anthropomorphic phantoms. However, for single in vivo measurements (<1.7 mGy) the sensitivity is too low.

Project description:PurposeVascular remodeling is a significant pathological feature of various pulmonary diseases, which may be assessed by quantitative computed tomography (CT) imaging. The purpose of this study was therefore to develop and validate an automatic method for quantifying pulmonary vascular morphology in CT images.MethodsThe proposed method consists of pulmonary vessel extraction and quantification. For extracting pulmonary vessels, a graph-cuts-based method is proposed which considers appearance (CT intensity) and shape (vesselness from a Hessian-based filter) features, and incorporates distance to the airways into the cost function to prevent false detection of airway walls. For quantifying the extracted pulmonary vessels, a radius histogram is generated by counting the occurrence of vessel radii, calculated from a distance transform-based method. Subsequently, two biomarkers, slope α and intercept β, are calculated by linear regression on the radius histogram. A public data set from the VESSEL12 challenge was used to independently evaluate the vessel extraction. The quantitative analysis method was validated using images of a three-dimensional (3D) printed vessel phantom, scanned by a clinical CT scanner and a micro-CT scanner (to obtain a gold standard). To confirm the association between imaging biomarkers and pulmonary function, 77 scleroderma patients were investigated with the proposed method.ResultsIn the independent evaluation with the public data set, our vessel segmentation method obtained an area under the receiver operating characteristic (ROC) curve of 0.976. The median radius difference between clinical and micro-CT scans of a 3D printed vessel phantom was 0.062 ± 0.020 mm, with interquartile range of 0.199 ± 0.050 mm. In the studied patient group, a significant correlation between diffusion capacity for carbon monoxide and the biomarkers, α (R = -0.27, P = 0.018) and β (R = 0.321, P = 0.004), was obtained.ConclusionIn conclusion, the proposed method was validated independently using a public data set resulting in an area under the ROC curve of 0.976 and using a 3D printed vessel phantom data set, showing a vessel sizing error of 0.062 mm (0.16 in-plane pixel units). The correlation between imaging biomarkers and diffusion capacity in a clinical data set confirmed an association between lung structure and function. This quantification of pulmonary vascular morphology may be helpful in understanding the pathophysiology of pulmonary vascular diseases.

Project description:Purpose To develop a method to incorporate the propagation of contrast material into computational anthropomorphic phantoms for estimation of organ dose at computed tomography (CT). Materials and Methods A patient-specific physiologically based pharmacokinetic (PBPK) model of the human cardiovascular system was incorporated into 58 extended cardiac-torso (XCAT) patient phantoms. The PBPK model comprised compartmental models of vessels and organs unique to each XCAT model. For typical injection protocols, the dynamics of the contrast material in the body were described according to a series of patient-specific iodine mass-balance differential equations, the solutions to which provided the contrast material concentration time curves for each compartment. Each organ was assigned to a corresponding time-varying iodinated contrast agent to create the contrast material-enhanced five-dimensional XCAT models, in which the fifth dimension represents the dynamics of contrast material. To validate the accuracy of the models, simulated aortic and hepatic contrast-enhancement results throughout the models were compared with previously published clinical data by using the percentage of discrepancy in the mean, time to 90% peak, peak value, and slope of enhancement in a paired t test at the 95% significance level. Results The PBPK model allowed effective prediction of the time-varying concentration curves of various contrast material administrations in each organ for different patient models. The contrast-enhancement results were in agreement with results of previously published clinical data, with mean percentage, time to 90% peak, peak value, and slope of less than 10% (P > .74), 4%, 7%, and 14% for uniphasic and 12% (P > .56), 4%, 12%, and 14% for biphasic injection protocols, respectively. The exception was hepatic enhancement results calculated for a uniphasic injection protocol for which the discrepancy was less than 25%. Conclusion A technique to model the propagation of contrast material in XCAT human models was developed. The models with added contrast material propagation can be applied to simulate contrast-enhanced CT examinations. © RSNA, 2017 Online supplemental material is available for this article.

Project description:PurposeThe aim of this paper is to describe a public, open-access, computed tomography (CT) phantom image set acquired at three centers and collected especially for radiomics reproducibility research. The dataset is useful to test radiomic features reproducibility with respect to various parameters, such as acquisition settings, scanners, and reconstruction algorithms.Acquisition and validation methodsThree phantoms were scanned in three independent institutions. Images of the following phantoms were acquired: Catphan 700 and COPDGene Phantom II (Phantom Laboratory, Greenwich, NY, USA), and the Triple modality 3D Abdominal Phantom (CIRS, Norfolk, VA, USA). Data were collected at three Dutch medical centers: MAASTRO Clinic (Maastricht, NL), Radboud University Medical Center (Nijmegen, NL), and University Medical Center Groningen (Groningen, NL) with scanners from two different manufacturers Siemens Healthcare and Philips Healthcare. The following acquisition parameter were varied in the phantom scans: slice thickness, reconstruction kernels, and tube current.Data format and usage notesWe made the dataset publically available on the Dutch instance of "Extensible Neuroimaging Archive Toolkit-XNAT" (https://xnat.bmia.nl). The dataset is freely available and reusable with attribution (Creative Commons 3.0 license).Potential applicationsOur goal was to provide a findable, open-access, annotated, and reusable CT phantom dataset for radiomics reproducibility studies. Reproducibility testing and harmonization are fundamental requirements for wide generalizability of radiomics-based clinical prediction models. It is highly desirable to include only reproducible features into models, to be more assured of external validity across hitherto unseen contexts. In this view, phantom data from different centers represent a valuable source of information to exclude CT radiomic features that may already be unstable with respect to simplified structures and tightly controlled scan settings. The intended extension of our shared dataset is to include other modalities and phantoms with more realistic lesion simulations.

Dataset Information

Automatic Synthesis of Anthropomorphic Pulmonary CT Phantoms.

Publications

Automatic Synthesis of Anthropomorphic Pulmonary CT Phantoms.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets