Abstract
Background/Aim: To evaluate the impact of high data rate and sampling frequency detector technology compared to standard scan equipment on the image quality in abdominal computed tomography (CT) of overweight and obese patients. Patients and Methods: A total of 173 patients were retrospectively included in this study. Objective image quality in abdominal CT was evaluated using comparative analysis with new detector technology prior to market launch and standard CT equipment. Contrast noise ratio (CNR), image noise, volumetric computed tomography dose index (CTDIVol), and figures of merit (Q and Q1) were assessed for all patients. Results: Image quality was superior in the new detector technology for all parameters evaluated. The dose dependent parameters Q and Q1 showed a significant difference (p<0.001). Conclusion: A significant increase in objective image quality could be demonstrated using a new generation detector setup with increased frequency transfer in abdominal CT of overweight patients.
Overweight and obesity are factors increasingly impacting the image quality in diagnostic radiology. It is described as a pandemic with a continuously increasing prevalence. Recent estimates suggest that overweight and obesity cause annually more than 1.2 million deaths in Europe. Additionally, obesity is the fourth cause of mortality after high blood pressure and dietary risks such as metabolic syndrome and tobacco consumption, corresponding to 13% of total deaths (1). Obesity is linked to disproportionality of body weight and height, which is reflected by the body mass index (BMI) (2). Early European studies have indicated a rise in the prevalence of overweight and obesity, or mean BMI, also in the younger population during the COVID-19 pandemic (3-5). According to the World Health Organization (WHO) criteria (6), a body mass index (BMI) of 18.5-24.9 kg/m2 is considered normal weight, while a BMI of 25-29.9 kg/m2 is considered overweight or pre-obese. A BMI ≥30 kg/m2 is defined as obese, and the obese category can be sub-divided into obese class I (30-34.9 kg/m2), obese class II (35-39.9 kg/m2), and obese class III (≥40 kg/m2).
It is a constant challenge in radiology to perform imaging studies with acceptable diagnostic quality in overweight and obese patients, as they are commonly degraded by increased image noise and artefacts, i.e., beam hardening and photon starvation (7-9). This hampers image interpretation in daily routine (10). There is a well-known correlation between diagnostic irradiation of CT and the small but increased cancer risk, especially in the younger population (11-13). Thus, it is mandatory to optimize the radiation dose and image quality in obese patients.
Technically, CT detectors have a limited sampling or reading rate, that is, the speed at which data can be processed and transferred from the detector to the image reconstruction system.
At a given sampling rate, the amount of data, i.e., the number of projections available, for a dedicated rotation decreases for faster rotation times. For example, with the traditional sampling rate of 4kHz, a detector can process approximately 2,000 projections in one rotation when operating with a rotation time of 0.5 s, while the number of projections reduces to 1,000 in one rotation when operating at a rotation time of 0.25 s. This influences the (azimuthal) image resolution especially in regions away from the isocenter, where the overlap of projections is smaller, with higher reading sampling allowing for better resolution. One vendor recently introduced a CT scanner that offers a fast gantry rotation time of 0.25 s with scan speeds of up to 261 mm/s and a detector that has been designed with a sampling rate of up to 8 kHz to maintain image quality and spatial resolution even for large patients and across the whole field of view.
Patients and Methods
Patient population. This study was approved by the local Ethics Committee (EKNZ Nr. 2021-02299).
This study was performed on a prototype CT scanner prior to market launch that was equipped with a novel data transmission technology and higher maximum detector sampling frequency and larger gantry (SOMATOM X.ceed, Siemens Healthineers, Forchheim, Germany).
Consecutive patients imaged with this dedicated equipment between 02/2021 and 11/2021 were retrospectively searched (scan A). Patients that declined or could not give general consent to use of their data for research purposes were excluded from the study. Only adult patients with a contrast-enhanced CT scan of the abdomen and that had prior CT imaging of the abdomen at our department on a second-generation dual-source CT (SOMATOM Definition Flash, Siemens Healthineers, Forchheim, Germany) (scan B) were attributable for comparative analysis.
The patients underwent CT examinations for various reasons using dedicated standardized clinical scan protocols. Automatic tube current modulation and automatic tube voltage selection were activated for all acquisitions (scan A and B). The contrast injection protocol was the same for all patients. Patient demographics, BMI, scan parameters and dosimetry values were collected for every CT scan. All images were reconstructed using a slice thickness and increment of 3 mm.
Scan protocol. All patients underwent standardized routine scans of the abdomen in the context of their clinical workup for various reasons. The scan protocol and the injection protocol were predefined for all patients. Scan setup A was used with scan protocol as follows: 120 reference kilo Voltage (refkV), 95 basic quality (BQ) level, pitch 0.8, rotation time 0.5 seconds (s), 128×0.6 mm collimation and 3/3 mm reconstruction using Br40-kernel. Scan setup B was used with scan protocol as follows: 120 refkV, 280 reference milliampere seconds (mAs), pitch 1.2, rotation time 0.5 s, 128×0.6 mm collimation, 3/3 mm reconstruction using I31f-kernel. Since scan A used a new user interface including newly introduced image quality reference predefinitions, scan parameters were adapted according to the manufacturer to maintain the same radiation dose for all patients. For both setups, contrast injection consisted of 80 ml Iopamidol 370 mg/ml with 30 ml saline chaser bolus with a flow rate of 3 ml/s and a delay of 60 s.
Image quality analysis. Image analysis was performed using Picture Archiving and Communications Systems (GE Centricity version I6, GE Healthcare, Chalfont, St Giles, UK). Two readers (17 years and 3 years of experience in CT imaging) evaluated the images in a consensus reading. Readers were free to adjust zoom factor, window level and window width.
Comparative assessment of objective image quality was performed according to previously established methods (14). Radiological image quality is usually evaluated using noise, contrast, and spatial resolution. To assess the whole CT image acquisition process, one must also consider specific CT acquisition parameters, such as slice thickness and CT dose index (CTDI). Volumetric computed tomography dose index (CTDIVol) is a well-known measure of radiation dose in CT (15).
Using Rose theory, it is possible to build a figure of merit (Q) that comprises several quantities for assessment in a single quantitative index. These may be defined and presented in different ways and include the Contrast-to-Noise-Ratio (CNR) and a dosimetric quantity (CTDIVol) (14).
Q was defined as 
,
as previously described (14, 16), where CNR is the contrast to noise ratio. Contrast was evaluated as the difference between mean CT values in a standardized region of interest (ROI), measured in the aortic lumen at the anatomical level of the hepatic vein confluence with the inferior vena cava and in the anterior subcutaneous fat at the same anatomical level (17, 18). When the SD of the aorta was measured, calcification and plaque on the aortic wall were avoided. Noise was the mean noise in the considered details, evaluated as standard deviation in the same ROIs. For all measurements, the size, shape, and position of the ROIs were kept constant among the two scan setups evaluated. Higher Q values indicate higher image quality relative to radiation dose.
Q1 is another figure of merit described for assessment of objective image quality as proposed before. It includes image noise, spatial resolution, slice thickness and radiation dose. Q1 was evaluated as follows:
where σ was defined as image noise, z as slice thickness and MTF as average modulation transfer function (14).
Statistics. Statistical analysis was performed using SPSS v.28.0.1.1 (SPSS Inc., Chicago, IL, USA). Data was tested for normality using Kolomogorov-Smirnov test, and the Student’s t-test for independent data was performed for the analysis of normally distributed continuous data. The Mann-Whitney U-test was used for non-normally distributed continuous data. Statistical significance was set at p<0.05.
Results
Patient population. A total of 410 patients (scan setup A) that met the inclusion criteria were scanned in the defined time-interval and could be considered for the project (m=222; f=188). Out of this group, 88 patients declined general consent for data use and were excluded. Of the remaining 322 patients, 123 met the WHO criteria of overweight (BMI ≥25 kg/m2) and 50 met the criteria of obesity (BMI≥30 kg/m2) and could be included in the analysis. Prior computed tomography examinations of the abdomen (scan setup B) could be found for 173 patients (mean age=72.3±12.9 years, m=104, f=69) that were part of the final analysis (Figure 1).
Study selection flow diagram.
Objective image quality. For all overweight patients, CNR and Q1 were normally distributed as assessed by the Kolmogorow-Smirnow test (p>0.05) for both scan setups, while all other parameters were not normally distributed (p<0.05). The mean objective image quality measure CNR was significantly higher for setup A (Figure 2); it was the same for Q and Q1 (Figure 3), while the radiation dose (CTDIVol) was also higher for setup A without showing significance (Figure 2). Image noise (σ) was nearly the same for both setups (Figure 2).
Boxplots demonstrating the distribution of objective image quality parameters for all patients. Setup A: blue, setup B: orange. Volumetric computed tomography dose index (CTDIVol), image noise (σ), contrast to noise ratio (CNR). CNR was significantly higher in setup A, CTDIVol and σ showed no significant difference.
Boxplots demonstrating distribution of figures of merit Q and Q1 for all patients; Q1 was scaled with *100 for comparison.
In the subgroup analysis of obese patients, only the image noise σ and Q1 were normally distributed, as assessed by the Kolmogorow-Smirnow test (p>0.05), while all other parameters showed no normal distribution (p<0.05). CNR, Q, Q1 and radiation dose were significantly higher for setup A. Image noise was lower in setup A but showed no significance (Table I).
Detailed results for objective image quality assessment.
Discussion
The detrimental effect of reduced image quality in CT of overweight or obese patients for the quality of interpretation in radiology reports is well known (10). Photon attenuation increases exponentially with increasing patient thickness. Hence, the aim in imaging obese patients must be to improve image quality according to the as low as reasonable achievable (ALARA) statement (19, 20). While obese patients are often subjected to repeated projections/scans and therefore increased radiation doses, despite the use of dose modulation techniques, acceptance of decreased image quality may result in a quasi-adherence to ALARA principles (21).
Over the last decade, several technological developments were introduced in CT to reduce the radiation dose while maintaining image quality, but mainly focusing on software-based modifications (18, 22, 23).
The data analyzed show the superiority of the scan setup with high-capacity data transmission technology and larger gantry in the assessment of objective image quality using dose dependent evaluation parameters in obese patients. In bariatric CT imaging cropping artefacts occur when portions of the patient fall outside the reconstruction field of view if it is selected to be smaller than the scan field of view (24). High frequency detectors allow for the reduction of such artefacts especially in regions away from the isocenter, where the overlap of projections is smaller, while allowing for increased image quality at the same time (25). Since obese patients often need the extent of most of the field of view to the peripheral parts, they might benefit from a scan setup with higher sampling rate that yields better image quality in the outer parts of the scan field.
Our study had several limitations. Firstly, the new scan setup (A) prior to market launch uses new AI-based tools for patient-specific dose modulation that are different to the other scan setup (B) used, even if provided by the same manufacturer. Therefore, the dose differed between the two groups, even if not significantly for all patients. Similar consecutive effects could be found for image noise, but less important. Both parameters for the objective assessment of image quality Q and Q1 respect radiation dose and noise. An increase in radiation dose and decrease of image noise both may have detrimental effects on Q and Q1. Yet both figures of merit Q and Q1 were significantly higher in scan A, indicating the superiority of this setup. Secondly, patients were included under the condition of any prior abdominal CT examination in the past with the dedicated setup to allow for comparison, so body weight might have changed between the scan intervals. Thirdly, aiming for standardization, we measured image quality parameters in standardized ROIs in the middle of the patient/gantry to allow direct comparison. But the major advantageous effect of higher sampling rate might affect the periphery of the scan field. We therefore aimed to define CNR as the difference between mean CT values measured in the aorta lumen and in the anterior subcutaneous fat at the same anatomical level that was more peripherally located. Moreover, the reconstruction kernels differed; this was due to the introduction of new kernels for scan A and prior kernels used for setup B were not in use any more on scan A. The choice of kernels for scan A was according to manufacturers’ advice for optimal translation.
Conclusion
In conclusion, our study could demonstrate a major impact on image quality in abdominal CT of overweight and obese patients, showing a significant increase in objective image quality of the new generation detector setup with increased frequency transfer. Since these patients may anatomically extend to the periphery of the field of view, they substantially benefit from increased image quality due to a higher data rate.
Footnotes
Authors’ Contributions
Conceptualization, FI, GC, HD, MS, EA, HLJ, KF, KRA, NA, NT; methodology, FI, GC, MS, EA, KF, NT; software, KF; formal analysis, FI, GC, MS, KRA, NT; resources, KF, NA, KRA; data curation, IF, GC, HD, NT; writing—original draft preparation, IF, NT; writing—review and editing, IF, MS, EA, KRA, NA, HL, NT; visualization, FI, NT; supervision, NT, KRA. All Authors have read and agreed to the published version of the manuscript.
Conflicts of Interest
IF received a scientific grant from Guerbet AG, Switzerland. FK is an employee of Siemens Healthcare GmbH. The remaining Authors declare that the research was conducted in the absence of any commercial or financial relationship that could be construed as a potential conflict of interest. No further funding or grant support was received for this study. Siemens Healthcare provided support in the form of salaries for author FK but did not have any additional role in the study design, data collection and analysis or decision to publish the manuscript.
- Received March 17, 2023.
- Revision received April 1, 2023.
- Accepted April 5, 2023.
- Copyright © 2023 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
