Diagnostic reference level quantities for adult chest and abdomen-pelvis CT examinations: correlation with organ doses

Objectives

To evaluate correlations between DRL quantities (DRLq) stratified into patient size groups for non-contrast chest and abdomen-pelvis CT examinations in adult patients and the corresponding organ doses.

Methods

This study presents correlations between DRLq  (CTDI_vol, DLP and SSDE) stratified into patient size ranges and corresponding organ doses shared in four groups: inside, peripheral, distributed and outside. The demographic, technical and dosimetric parameters were used to identify the influence of these quantities in organ doses. A robust statistical method was implemented in order to establish these correlations and its statistical significance.

Results

Median values of the grouped organ doses are presented according to the effective diameter ranges. Organ doses in the regions inside the imaged area are higher than the organ doses in peripheral, distributed and outside regions, excepted to the peripheral doses associated with chest examinations. Different levels of statistical significance between organ doses and the DRLq were presented.

Conclusions

Correlations between DRLq and target-organ doses associated with clinical practice can support guidance’s to the establishment of optimization criteria. SSDE demonstrated to be significant in the evaluation of organ doses is also highlighted. The proposed model allows the design of optimization actions with specific risk-reduction results.

1. DRL quantities and target-organ doses can support guidance’s to optimization criteria.
2. Organ doses inside imaged area are higher than in other regions.
3. Statistical significance between organ doses and the DRL quantities are presented.
4. Results reinforce the importance on the adequate choice of the scan length.

There is an increasing concern to implement optimized CT protocols, in order to adequate the radiation dose to the patient without loss of image quality and diagnostic information [1]. The diagnostic reference level (DRL) approach was introduced decades ago [2, 3] and adopted by International Commission on Radiological Protection (ICRP) in its publication 73 [4]. A complete guidance for the DRL practical application is presented in the ICRP 135 [5]. This publication emphasizes that the DRLs must adopt measurable quantities in order to assess and compare doses for particular types of examinations among different facilities and can be used as a tool to optimize medical exposures [6,7,8].

On the other side, the ICRP 147 publication [9] recognizes that the best way to estimate the risk to individuals submitted to medical procedures using ionizing radiation is the adoption of organ/tissue doses and specific dose-risk models. The correlation of these risks in low dose levels as that normally used in diagnostic imaging is difficult and associated with complex uncertainties. Therefore, connections between organ doses and measurable quantities determined using controlled cohorts and their associations with technical factors normally adopted in the clinical practice may support the development of models to identify patient-specific risks. For example, retrospective correlations between DRL quantities and target-organ doses associated with some clinical practice can support guidance’s in the establishment of optimization criteria. This kind of approach is in agreement with the American Association of Physicists in Medicine (AAPM) position related to Medical Imaging Radiation Limits [10].

The correlation between organ doses and dosimetric DICOM-reported information was recently addressed by AAPM and EFOMP [11]. In the case of CT examinations, SSDE has shown a significant correlation with organ doses [12], but still requires improvement in both its interpretation as a patient radiation protection predictor and its suitability as a risk-associated estimator in clinical situations where organs are partially irradiated. Although these limitations add uncertainties to the organ dose estimations from SSDE, the combination of accurate clinically relevant information into Monte Carlo methods associated with validated SSDE calculations represents a valuable effort in the direction of associate DICOM header and Radiation Dose Structured Report (RDSR) available information with organ doses and patient-risk models [11]. The current improvements of this correlation between SSDE and patient-risk models associated with the method proposed in the present work allow to quantify the dose reduction expected to a given patient cohort when an optimized protocol will be introduced. This kind of quantification is not possible using CTDI-based only methods.

The implementation of the proposed method may allow the association of different operational parameters of a CT scanner for a given clinical task and image quality with patient organ doses and consequent risk-related information. It can impact the clinical operation if adopted as a decision-making tool for comparing protocols in a commissioning or optimization task.

This study presents correlations between DRL quantities (CTDI_vol, DLP and SSDE) [5] stratified into adult patient size ranges for non-contrast chest and abdomen-pelvis CT examinations performed at the Institute of Radiology of the Medical School of the University of São Paulo (INRAD-HCFMUSP) and their corresponding organ doses shared in four groups associated with the imaging area: inside, peripheral, distributed and outside. The demographic, technical and dosimetric quantities were used to calculate organ doses in order to identify the influence of these quantities and doses in critical organs. A robust statistical method was implemented in order to investigate associations between these parameters.

Database management, patient cohort and dose-related quantities

Data from chest and abdomen-pelvis CT examinations were retrospectively registered between January and September 2020 at INRAD-HCFMUSP. The study was approved by the HCFMUSP Institutional Review Board^Footnote1 and informed consent was not required. The study included examinations performed on adult patients (> 18 years old) using four CT scanners: SCANNER 1 – Discovery 750 HD (GE Healthcare); SCANNER 2 – Aquilion CXL (Canon Medical Systems); SCANNER 3 and SCANNER 4 – Brilliance 64 (Philips Medical Systems).

Teamplay® platform (Siemens Healthineers) was used to select DICOM diagnostically validated image series corresponding to the cohorts of interest for the present study and their corresponding RDSR. These studies were sent to the institutional PACS (IntelliSpace PACS-Enterprise, Philips) and eFilm Workstation 3.1 (Merge Healthcare) was used to select the images of interest in each series. Patient demographic and dose data were extracted from the DICOM Header using ImageJ® (U. S. National Institutes of Health, Bethesda, Maryland, USA). Table 1 summarizes the basic acquisition and reconstruction parameters of the studied protocols. Additionally, the following demographic and technical parameters were also registered for each patient/protocol: gender, age, weight, height, voltage, pitch, collimation, TCM mode and mAs at central slice. Table 2 presents the number of collected examinations by each CT scanner and patient gender, age, weight, height and BMI, as well as the respective medians and minimum–maximum values. To be consistent with ICRP 135 [5] sample sizes for DRL evaluation, effective diameter (�) groups with less than 30 patients were excluded.