Dataset Information

Pseudo dual-energy CT-derived iodine mapping using single-energy CT data based on a convolution neural network.

ABSTRACT:

Objective

The objectives of this study are: (1) to develop a convolutional neural network model that yields pseudo high-energy CT (CT_{pseudo_high}) from simple image processed low-energy CT (CT_low) images, and (2) to create a pseudo iodine map (IM_pseudo) and pseudo virtual non-contrast (VNC_pseudo) images for thoracic and abdominal regions.

Methods

Eighty patients who underwent dual-energy CT (DECT) examinations were enrolled. The data obtained from 55, 5, and 20 patients were used for training, validation, and testing, respectively. The ResUnet model was used for image generation model and was trained using CT_low and high-energy CT (CT_high) images. The proposed model performance was evaluated by calculating the CT values, image noise, mean absolute errors (MAEs), and histogram intersections (HIs).

Results

The mean difference in the CT values between CT_{pseudo_high} and CT_high images were less than 6 Hounsfield unit (HU) for all evaluating patients. The image noise of CT_{pseudo_high} was significantly lower than that of CT_high. The mean MAEs was less than 15 HU, and HIs were almost 1.000 for all the patients. The evaluation metrics of IM and VNC exhibited the same tendency as that of the comparison between CT_{pseudo_high} and CT_high images.

Conclusions

Our results indicated that the proposed model enables to obtain the DECT images and material-specific images from only single-energy CT images.

Advances in knowledges

We constructed the CNN-based model which can generate pseudo DECT image and DECT-derived material-specific image using only simple image-processed CT_low images for the thoracic and abdominal regions.

SUBMITTER: Yuasa Y

PROVIDER: S-EPMC10630979 | biostudies-literature | 2023

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Pseudo dual-energy CT-derived iodine mapping using single-energy CT data based on a convolution neural network.

Yuasa Yuki Y Shiinoki Takehiro T Fujimoto Koya K Tanaka Hidekazu H

BJR open 20231018 1

<h4>Objective</h4>The objectives of this study are: (1) to develop a convolutional neural network model that yields pseudo high-energy CT (CT<sub>pseudo_high</sub>) from simple image processed low-energy CT (CT<sub>low</sub>) images, and (2) to create a pseudo iodine map (IM<sub>pseudo</sub>) and pseudo virtual non-contrast (VNC<sub>pseudo</sub>) images for thoracic and abdominal regions.<h4>Methods</h4>Eighty patients who underwent dual-energy CT (DECT) examinations were enrolled. The data obta ...[more]

PMID: 37942496

Similar Datasets

Project description:Background Assessments of the quantitative limitations among the six commercially available dual-energy (DE) CT acquisition schemes used by major CT manufacturers could aid researchers looking to use iodine quantification as an imaging biomarker. Purpose To determine the limits of detection and quantification of DE CT in phantoms by comparing rapid peak kilovoltage switching, dual-source, split-filter, and dual-layer detector systems in six different scanners. Materials and Methods Seven 50-mL iohexol solutions were used, with concentrations of 0.03-2.0 mg iodine per milliliter. The solutions and water sample were scanned five times each in two phantoms (small, 20-cm diameter; large, 30 × 40-cm diameter) with six DE CT systems and a total of 10 peak kilovoltage settings or combinations. Iodine maps were created, and the mean iodine signal in each sample was recorded. The limit of blank (LOB) was defined as the upper limit of the 95% confidence interval of the water sample. The limit of detection (LOD) was defined as the concentration with a 95% chance of having a signal above the LOB. The limit of quantification (LOQ) was defined as the lowest concentration where the coefficient of variation was less than 20%. Results The LOD range was 0.021-0.26 mg/mL in the small phantom and 0.026-0.55 mg/mL in the large phantom. The LOQ range was 0.07-0.50 mg/mL in the small phantom and 0.20-1.0 mg/mL in the large phantom. The dual-source and rapid peak kilovoltage switching systems had the lowest LODs, and the dual-layer detector systems had the highest LODs. Conclusion The iodine limit of detection using dual-energy CT systems varied with scanner and phantom size, but all systems depicted iodine in the small and large phantoms at or below 0.3 and 0.5 mg/mL, respectively, and enabled quantification at concentrations of 0.5 and 1.0 mg/mL, respectively. © RSNA, 2019 Online supplemental material is available for this article. See also the editorial by Hindman in this issue.

Project description:PurposeThe dual-energy CT (DECT) LiverVNC application class in the Siemens Syngo.via software has been used to perform non-iodine material decompositions. However, the LiverVNC application is designed with an optional size-specific calibration based on iodine measurements. This work investigates the effects of this iodine-based size-specific calibration on non-iodine material decomposition and benchmarks alternative methods for size-specific calibrations.MethodsCalcium quantification was performed with split-filter and sequential-scanning DECT techniques on the Siemens SOMATOM Definition Edge CT scanner. Images were acquired of the Gammex MECT abdomen and head phantom containing calcium inserts with concentrations ranging from 50-300 mgCa/ml. Several workflows were explored investigating the effects of size-specific dual-energy ratios (DERs) and the beam hardening correction (BHC) function in the LiverVNC application. Effects of image noise were also investigated by varying CTDIvol and using iterative reconstruction (ADMIRE).ResultsWith the default BHC activated, Syngo.via underestimated the calcium concentrations in the abdomen for sequential-scanning acquisitions, leaving residual calcium in the virtual non-contrast images and underestimating calcium in the enhancement images for all DERs. Activation of the BHC with split-filter images resulted in a calcium over- or underestimation depending on the DER. With the BHC inactivated, the use of a single DER led to an under- or overestimate of calcium concentration depending on phantom size and DECT modality. Optimal results were found with BHC inactivated using size-specific DERs. CTDIvol levels and ADMIRE had no significant effect on results.ConclusionWhen performing non-iodine material decomposition in the LiverVNC application class, it is important to understand the implications of the BHC function and to account for patient size appropriately. The BHC in the LiverVNC application is specific to iodine and leads to inaccurate quantification of other materials. The inaccuracies can be overcome by deactivating the BHC function and using size-specific DERs, which provided the most accurate calcium quantification.

Project description:ObjectivesTo quantitatively evaluate the impact of virtual monochromatic images (VMI) on reduced-iodine-dose dual-energy coronary computed tomography angiography (CCTA) in terms of coronary lumen segmentation in vitro, and secondly to assess the image quality in vivo, compared with conventional CT obtained with regular iodine dose.Materials and methodsA phantom simulating regular and reduced iodine injection was used to determine the accuracy and precision of lumen area segmentation for various VMI energy levels. We retrospectively included 203 patients from December 2017 to August 2018 (mean age, 51.7 ± 16.8 years) who underwent CCTA using either standard (group A, n = 103) or reduced (group B, n = 100) iodine doses. Conventional images (group A) were qualitatively and quantitatively compared with 55-keV VMI (group B). We recorded the location of venous catheters.ResultsIn vitro, VMI outperformed conventional CT, with a segmentation accuracy of 0.998 vs. 1.684 mm2, respectively (p < 0.001), and a precision of 0.982 vs. 1.229 mm2, respectively (p < 0.001), in simulated overweight adult subjects. In vivo, the rate of diagnostic CCTA in groups A and B was 88.4% (n = 91/103) vs. 89% (n = 89/100), respectively, and noninferiority of protocol B was inferred. Contrast-to-noise ratios (CNR) of lumen versus fat and muscle were higher in group B (p < 0.001) and comparable for lumen versus calcium (p = 0.423). Venous catheters were more often placed on the forearm or hand in group B (p < 0.001).ConclusionIn vitro, low-keV VMI improve vessel area segmentation. In vivo, low-keV VMI allows for a 40% iodine dose and injection rate reduction while maintaining diagnostic image quality and improves the CNR between lumen versus fat and muscle.Key points• Dual-energy coronary CT angiography is becoming increasingly available and might help improve patient management. • Compared with regular-iodine-dose coronary CT angiography, reduced-iodine-dose dual-energy CT with low-keV monochromatic image reconstructions performed better in phantom-based vessel cross-sectional segmentation and proved to be noninferior in vivo. • Patients receiving reduced-iodine-dose dual-energy coronary CT angiography often had the venous catheter placed on the forearm or wrist without compromising image quality.

Dataset Information

Pseudo dual-energy CT-derived iodine mapping using single-energy CT data based on a convolution neural network.

Objective

Methods

Results

Conclusions

Advances in knowledges

Publications

Pseudo dual-energy CT-derived iodine mapping using single-energy CT data based on a convolution neural network.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets