Assessment of thyroid iodine accumulation following repeated iodinated contrast media administration using dual-energy computed tomography in a rabbit model

Abstract

Repeated exposure to iodinated contrast media (ICM) may induce thyroid dysfunction but supporting radiological and histological evidence remains limited. This study aimed to evaluate whether repeated ICM administration induces iodine accumulation and related functional and histological changes in the thyroid. This prospective 8-week study involved 30 male New Zealand white rabbits (0.8–1.2 kg, < 1 year) assigned to four groups. The control group (n = 8) received 0.9% sodium chloride solution, while experimental groups received intravenous iohexol three times weekly at 1.6 mL/kg (n = 7), 6.0 mL/kg (n = 7), or 12.0 mL/kg (n = 8). Non-contrast thyroid dual-energy computed tomography (DECT) was performed weekly to quantify iodine accumulation, expressed as the thyroid enhancement ratio (TER) relative to paraspinal muscle. Serum free thyroxine (fT4), free triiodothyronine (fT3), thyroid-stimulating hormone, and creatinine were measured at baseline, week 4, and week 8. Histological analysis was performed at week 8. At week 8, TER was significantly higher in the 6.0 mL/kg (2.59 ± 0.89) and 12.0 mL/kg (2.79 ± 0.77) groups than in controls (1.15 ± 0.42) and the 1.6 mL/kg group (1.32 ± 0.46) (all P < 0.05). While transient fT4 and fT3 elevations occurred and nodular hyperplasia, follicular atrophy, and inflammation were more frequent in high-dose groups, neither showed statistically significant intergroup differences (both P > 0.05). In conclusion, repeated intravenous administration of high-dose ICM in rabbits was associated with increased iodine accumulation in the thyroid gland, as quantified by DECT. However, a transient rise in serum thyroid hormone levels, as well as more frequent histopathological changes, did not reach statistical significance.

Introduction

Radiographic contrast media can enhance the visualization of internal organs and structures during diagnostic imaging procedures¹. Computed tomography (CT) contrast agents generally contain iodine-based compounds, most commonly featuring a chemically modified 2,4,6-tri-iodinated benzene ring². Although iodinated contrast media (ICM) substantially improve diagnostic accuracy, their use is associated with adverse effects ranging from mild allergic reactions to severe anaphylaxis, contrast-induced nephropathy, and thyroid dysfunction³.

In 2022, the U.S. Food and Drug Administration recommended thyroid function testing in neonates and children aged 3 years or younger following intravascular exposure to ICM⁴. This guideline was subsequently updated to incorporate individual risk factors—such as prematurity, low birth weight, and preexisting thyroid disease—due to limited evidence linking ICM exposure to clinically significant hypothyroidism⁵. Previous studies have reported a variable incidence of thyroid dysfunction following contrast-enhanced CT, ranging from 1.0% to 14.0%, but most cases were classified as transient subclinical hypothyroidism^6–8. Higher doses of ICM, such as those administered during cardiac catheterization, have been associated with increased incidence rates (2.9–28.6%), yet overt hypothyroidism remains rare even after repeated or high-dose exposure^6,9–11. These hormonal alterations are believed to result from the Wolff–Chaikoff effect, a temporary suppression of thyroid hormone synthesis caused by intrathyroidal iodide overload⁶. Nevertheless, direct evidence of intrathyroidal iodine accumulation or corresponding histopathologic changes following ICM exposure has not been clearly demonstrated.

Recently, dual-energy computed tomography (DECT) has enabled spectral material decomposition and differentiation of substances such as water, iodine, calcium, and uric acid by estimating effective atomic numbers^12–14. We hypothesized that non-contrast DECT iodine mapping could enable in vivo assessment of iodine accumulation in the thyroid gland.

The purpose of this experimental study is to evaluate iodine accumulation in the thyroid gland using DECT imaging and to assess associated thyroid function and histopathologic changes following repeated ICM administration in a rabbit model.

Materials and methods

Statement

This study was approved by the Institutional Animal Care and Use Committee of Seoul National University Hospital (IACUC No. 21-0230-C1A1). Experiments were conducted in strict accordance with institutional and international guidelines for the care and use of laboratory animals, and the study adhered to the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines. For euthanasia, animals under deep anesthesia received an intravenous administration of an excessive dose (7–10 ml) of 2% xylazine hydrochloride (2 mg/kg, Rompun; Bayer-Schering Pharma, Republic of Korea).

Study timeline

Thirty male New Zealand white rabbits (0.8–1.2 kg, < 1 year old) were obtained from Koatech (Pyeongtaek, Korea) and maintained under standard laboratory conditions with free access to food and water. The rabbits were assigned to four groups: the control group (n = 8) received intravenous 0.9% sodium chloride solution, while the experimental groups received iodinated contrast media (iohexol, 300 mg•I/mL; Encormac, Seoul, Republic of Korea) at doses of 1.6 mL/kg (n = 7), 6.0 mL/kg (n = 7), or 12.0 mL/kg (n = 8), administered three times per week (Fig. 1). The 1.6 mL/kg dose simulated a clinical dose, while the 6.0 mL/kg and 12.0 mL/kg doses exceeded the clinical use range¹⁵. No iodine-restricted diet was used, and water was provided ad libitum. Before procedures and imaging, rabbits were anesthetized with an intravenous injection of zolazepam (5 mg/kg, Zoletil; Virbac, France) and 2% xylazine hydrochloride (2 mg/kg, Rompun).

Fig. 1.

Study flow diagram. The control group (n = 8) received 0.9% sodium chloride solution, while the experimental groups received iodinated contrast media (iohexol, 300 mg•I/mL) at doses of 1.6 mL/kg (n = 7), 6.0 mL/kg (n = 7), and 12.0 mL/kg (n = 8), administered three times per week. Dual-energy computed tomography (DECT) scans were performed weekly before iodinated contrast media injections, totaling eight scans per rabbit. Serum thyroid function test (TFT) were conducted at baseline, week 4, and week 8. At study completion, thyroid glands were harvested and stained with hematoxylin and eosin for histopathological evaluation. DECT Dual-energy computed tomography, TFT thyroid function test.

DECT scans were performed weekly before ICM injections, totaling eight scans per rabbit, and serum thyroid function tests at baseline, week 4, and week 8. After completing the study, rabbits were sacrificed, and their thyroid glands were harvested and stained with hematoxylin and eosin (H&E) for histopathological evaluation.

Evaluation of iodine accumulation

All rabbits underwent non-contrast thyroid DECT using a 256-row CT scanner (Revolution Apex; GE Healthcare, Chicago, IL). The DECT parameters were 80 kVp/250 mAs and 140 kVp/145 mAs, with a gantry rotation time of 0.5 s, a slice thickness of 1.25 mm at 1.25 mm intervals, a pitch of 0.984, and detector collimation of 0.625 mm × 64. Virtual monoenergetic imaging (VMI) was used to reconstruct images at 40, 50, and 70 keV. CT images were transferred to a post-processing workstation (GSI viewer; GE Healthcare, Chicago, IL) for measurement of iodine concentration in the thyroid glands. Image datasets obtained at different voltages for the same slices were analyzed with a slice thickness of 1.25 mm to assess iodine concentration.

Two radiologists (C.R., with 2 years of experience in radiology, and S.L., with 13 years of experience in pediatric radiology) measured iodine concentrations (mg/mL) in the thyroid gland, paraspinal muscle, and sternocleidomastoid muscle. Slices for the region of interest (ROI) were carefully selected to minimize beam hardening artifacts and ensure homogeneity. The thyroid ROIs were manually outlined on the right and left lobes using a freehand method (Fig. 2a). Muscle ROIs were marked as circular regions on both sides, each totaling at least 10 mm² (Fig. 2b). The attenuation values of the thyroid gland and muscles were defined as the mean attenuation within the ROI area.

Fig. 2.

Region of interest (ROI) selection for the thyroid gland, paraspinal muscles, and sternocleidomastoid muscles. ROIs for (a) the thyroid gland and (b) the paraspinal and sternocleidomastoid muscles were delineated on an iodine map obtained using dual-energy computed tomography in a rabbit administered iohexol at 12 mL/kg for 8 weeks. The total area and mean attenuation values were: (a) thyroid gland, 15.10 mm² and 20.70 HU; (b) paraspinal muscle, 32.64 mm² and 2.13 HU; and sternocleidomastoid muscle, 13.18 mm² and 3.68 HU. ROI region of interest, HU Hounsfield unit.

The thyroid enhancement ratio (TER) was calculated as the ratio of thyroid gland attenuation to paraspinal muscle attenuation, normalized to the baseline ratio, and averaged between the two radiologists¹⁶. TER was also computed using the sternocleidomastoid muscle to evaluate the reliability of the results.

Thyroid function test

Levels of free thyroxine (fT4), free triiodothyronine (fT3), and thyroid-stimulating hormone (TSH) were measured using an electrochemiluminescence immunoassay (Cobas^® e801 system; Roche Diagnostics, Mannheim, Germany). The reference ranges were 0.8–1.8 ng/dL for fT4, 2.3–4.2 pg/mL for fT3, and 0.4–4.0 mIU/L for TSH. Subclinical hypothyroidism was defined as a TSH level between 5.0 and 10.0 mIU/L with normal fT4 levels¹⁷. Serum creatinine levels were also assessed every 4 weeks. All values were calculated as ratios relative to baseline for comparison and plotting. Blood samples were collected before ICM administration.

Histopathology

At the end of week 8, thyroid tissues were harvested and stained with H&E for histological evaluation. Samples with inadequate staining or improper excision (e.g., including muscle tissue) were excluded. Thyroid tissue changes were categorized into nodular hyperplasia, follicular atrophy, and inflammation, which are associated with thyroid dysfunction. A pathologist (J.K., with 13 years of experience in pathology) evaluated each category as normal, mild, or moderate, assigning scores of 0, 1, and 2, respectively. The frequency and severity of histological changes were compared between the control and experimental groups.

Statistical analyses

Differences in values between groups were analyzed using the Kruskal–Wallis test, followed by Dunn’s test with Bonferroni correction for post hoc comparisons. The intraclass correlation coefficient (ICC) for TER values, calculated using the paraspinal and sternocleidomastoid muscles, was assessed with a two-way random-effects model. Inter-rater agreement was assessed using ICCs for thyroid gland and paraspinal muscle attenuation, based on measurements from two radiologists. ICC values were interpreted as follows: <0.50, poor; 0.50–0.75, moderate; 0.75–0.90, good; ≥0.90, excellent reliability¹⁸. Statistical analyses were performed using Python (v.3.8, Python Software Foundation, Beaverton, OR) with NumPy (v.1.25.0), Pandas (v.2.1.0), SciPy (v.1.11.0), Statsmodels (v.0.15.0), Matplotlib (v.3.7.2), and Seaborn (v.0.13.2).

Results

Quantification of iodine accumulation

Figure 3 illustrates weekly TER measurements assessed by DECT. The average ROI area for thyroid gland measurements was 23.5 ± 12.9 mm². The baseline iodine concentrations of the paraspinal muscle, sternocleidomastoid muscle, and thyroid gland were 2.48 ± 0.36 mg/mL, 3.02 ± 0.47 mg/mL, and 7.52 ± 0.56 mg/mL, respectively.

Fig. 3.

Time series of thyroid enhancement ratios (TERs). TERs over time in rabbits from the control group and experimental groups administered iohexol three times per week at doses of 1.6 mL/kg, 6.0 mL/kg, and 12.0 mL/kg. TER was calculated as the ratio of thyroid gland attenuation to paraspinal muscle attenuation, normalized to the baseline value. The 6.0 and 12.0 mL/kg groups exhibited significantly higher iodine accumulation compared to the control and 1.6 mL/kg groups. TER thyroid enhancement ratio.

At week 8, TER values were as follows: 1.15 ± 0.42 in the control group, 1.32 ± 0.46 in the 1.6 mL/kg group (P = 0.99 vs. control), 2.59 ± 0.89 in the 6.0 mL/kg group (P = 0.006 vs. control; P = 0.045 vs. 1.6 mL/kg), and 2.79 ± 0.77 in the 12.0 mL/kg group (P < 0.001 vs. control; P = 0.006 vs. 1.6 mL/kg; P = 0.99 vs. 6.0 mL/kg). The 6.0 and 12.0 mL/kg groups exhibited significantly higher iodine accumulation compared to the control and 1.6 mL/kg groups. Representative iodine maps from the 1.6 and 6.0 mL/kg groups are shown in Fig. 4. The ICC between TER values calculated using paraspinal and sternocleidomastoid muscles was 0.860, indicating good reliability. The inter-rater agreement for thyroid gland attenuation, paraspinal muscle attenuation, and TER were 0.732 and 0.761, and 0.802, respectively, indicating moderate to good reliability.

Fig. 4.

Iodine maps of the thyroid gland acquired by dual-energy computed tomography (DECT). Representative iodine maps obtained using DECT: (a) baseline and (b) after 8 weeks of iohexol administration at 1.6 mL/kg three times per week; (c) baseline and (d) after 8 weeks at 6.0 mL/kg. Progressive iodine accumulation is observed in the thyroid glands of rabbits that received higher doses of iodinated contrast media. DECT dual-energy computed tomography.

Thyroid function test

Table 1 presents the mean serum concentrations of fT4 and fT3, while Fig. 5 illustrates the ratios of fT4 and fT3 relative to baseline at 4 and 8 weeks in each group. In the 6.0 and 12.0 mL/kg groups, fT4 levels increased temporarily in week 4, followed by a decrease in week 8, compared to the control group. fT3 levels remained elevated at both time points in these groups. Baseline fT4 levels were within the normal range in all groups. At week 4, two subjects in both the 6.0 mL/kg and 12.0 mL/kg groups exceeded the upper normal limit (1.8 ng/dL); at week 8, one subject in the 12.0 mL/kg group exceeded this limit. Baseline fT3 levels were within the normal range in all groups. At week 4, two subjects in both the 6.0 mL/kg and 12.0 mL/kg groups exceeded the upper normal limit (4.2 pg/mL); at week 8, one subject in the 6.0 mL/kg group and two subjects in the 12.0 mL/kg group exceeded this limit. However, none of the differences in fT4 or fT3 ratios were statistically significant when comparing the control group with the 1.6, 6.0, and 12.0 mL/kg groups at week 8 (fT4, P = 0.231, 0.831, and 0.528; fT3, P = 0.902, 0.164, and 0.131, respectively). TSH levels were suppressed (< 0.01 mIU/L) in all animals throughout the study; therefore, comparison of the incidence of subclinical hypothyroidism was not feasible. Serum creatinine levels increased slightly in all groups, but without statistical significance (P = 0.670).

Table 1.

Serum concentrations of free thyroxine (fT4) and free triiodothyronine (fT3).

Group	fT4 (ng/dL)			fT3 (pg/mL)
	Baseline	Week 4	Week 8	Baseline	Week 4	Week 8
Control (n = 8)	1.375 (0.492)	1.344 (0.395)	1.373 (0.283)	3.863 (0.269)	3.773 (0.364)	3.843 (0.250)
1.6 mL/kg (n = 7)	1.179 (0.212)	1.419 (0.076)	1.564 (0.223)	3.735 (0.189)	3.758 (0.181)	3.759 (0.137)
6.0 mL/kg (n = 7)	1.378 (0.438)	1.678 (0.485)	1.524 (0.375)	3.739 (0.387)	3.987 (0.349)	3.992 (0.338)
12.0 mL/kg (n = 8)	1.376 (0.452)	1.691 (0.245)	1.530 (0.404)	3.669 (0.291)	3.822 (0.296)	3.908 (0.229)

Values represent the mean with standard deviation in parentheses. fT4 free thyroxine, fT3 free triiodothyronine.

Fig. 5.

Variations in thyroid hormone levels following repeated iodinated contrast media administration. Hormone levels are expressed as ratios relative to baseline values. (a) A transient increase of free thyroxine (fT4) was observed in week 4 in the 6.0 and 12.0 mL/kg groups, followed by a decline in week 8. (b) Elevated levels of free triiodothyronine (fT3) were sustained at both weeks 4 and 8 in the high-dose groups compared to controls. No statistically significant differences were found in fT4 or fT3 levels among the groups. fT4 free thyroxine, fT3 free triiodothyronine.

Histopathology

Table 2 summarizes histologic findings at week 8 following ICM administration. Samples available for evaluation included 7 in the control group, 6 in the 1.6 mL/kg group, 7 in the 6.0 mL/kg group, and 7 in the 12.0 mL/kg group. Three samples (one from each group except the 6.0 mL/kg group) were excluded due to inadequate preservation or staining. Representative gross and histopathologic images from the 6.0 mL/kg group are presented in Fig. 6, showing findings such as nodular hyperplasia, follicular atrophy, and inflammation. Although these changes were more frequently observed at higher ICM doses, the differences among groups did not reach statistical significance (control vs. 1.6 mL/kg group, P = 0.99; control vs. 6.0 mL/kg group, P = 0.108; control vs. 12.0 mL/kg group, P = 0.134).

Table 2.

Histopathological abnormality scores.

Group	Nodular hyperplasia	Follicular atrophy	Inflammation
Control group (n = 7)	0.14 (0.38)	0.00 (0.00)	0.14 (0.38)
1.6 mL/kg (n = 6)	0.00 (0.00)	0.33 (0.52)	0.33 (0.52)
6.0 mL/kg (n = 7)	0.71 (0.49)	0.71 (0.76)	0.57 (0.53)
12.0 mL/kg (n = 7)	0.57 (0.79)	0.57 (0.53)	0.71 (0.49)

All values represent the mean severity score, with the standard deviation in parentheses. Findings were classified as normal (0), mild (1), or moderate (2).

Fig. 6.

Gross and histopathological findings of the thyroid gland in the 6.0 mL/kg group. (a) Gross image of the excised thyroid gland from a rabbit after 8 weeks of repeated iodinated contrast media administration at 6.0 mL/kg, three times per week (scale bar, 10 mm). (b–d) Hematoxylin and eosin-stained histological sections of the thyroid (scale bar, 300 μm) demonstrate typical abnormalities observed in the high-dose group: (b) nodular hyperplasia, (c) follicular atrophy, and (d) inflammation.

Discussion

In this study, repeated intravenous administration of high-dose iodinated contrast media in rabbits was associated with increased iodine accumulation in the thyroid gland, as quantified by DECT. However, the incidence of subclinical hypothyroidism could not be assessed, and the between-group histopathological analysis did not reach statistical significance, limiting the clinical interpretation of the findings.

Previous studies have explored whether ICM induces hypothyroidism in pediatric patients, leading to ongoing debate regarding the necessity of thyroid function monitoring following ICM administration^4–8. Identified risk factors include patient characteristics such as age under 3 months, congenital heart disease, and impaired renal function^5,10,19, as well as procedural variables including cumulative ICM dose^10,20. A previous study identified a cumulative ICM dose of 8.7 g/kg (equivalent to 29 mL/kg for a 300 mg•I/mL concentration) as a potential predictor of subclinical hypothyroidism, particularly in cases of extremely high cumulative ICM doses resulting from cardiac procedures²⁰. However, another study reported no statistically significant difference between infants with hypothyroidism and euthyroid infants, with cumulative ICM doses of 3.6 mL/kg and 3.3 mL/kg, respectively, suggesting no clear association with hypothyroidism¹⁰. Therefore, both clinically relevant ICM dose (1.6 mL/kg) and extremely high doses (6.0 and 12.0 mL/kg) were intentionally administered to exaggerate potential effects and facilitate the detection of iodine accumulation.

Recent advances in DECT enable material decomposition and iodine-specific mapping, facilitating the assessment of iodine distribution in various tissues^13,21,22. Most previous DECT studies have focused on contrast-enhanced imaging for tumor detection and characterization^21–23; however, non-contrast DECT also enables the quantification of native iodine content, including in thyroid tissue. The DECT system employed in this study has been previously validated for accurate iodine quantification^13,24,25. Although the TER used in this study has not been thoroughly validated, and using muscle as a reference may introduce variability due to factors such as hydration or perfusion differences, minimal baseline variability was observed, with good interobserver agreement when using both the paraspinal and sternocleidomastoid muscles as references. However, technical factors such as tube voltage, VMI energy level, and detectable iodine dose can still affect the accuracy of quantitative DECT measurements^26,27. In this study, iodine mapping was performed using a single fast-kVp switching DECT system under standardized acquisition parameters, and all measured values in this study exceeded the detection threshold of 0.4–1.5 mg I/mL²⁸.

Although our results indicate that repetitive high-dose iodinated contrast media exposure resulted in significant intrathyroidal iodine accumulation, which could theoretically affect thyroid hormone synthesis, we were unable to adequately demonstrate this effect in the present study. We hypothesized that excess intrathyroidal iodine accumulation triggers the Wolff–Chaikoff effect, resulting in transient inhibition of thyroid hormone synthesis and subsequent ICM–induced hypothyroidism²⁹. Failure to escape this inhibitory response may lead to persistent hypothyroidism³⁰. Previous studies in hyperthyroid patients demonstrated a negative correlation between DECT-derived iodine concentration and radioactive iodine uptake, reflecting rapid iodine turnover in hyperfunctioning glands^23,31. However, evaluation of subclinical hypothyroidism was limited by suppressed TSH levels (< 0.01 mIU/L), which are essential for assessing hypothalamic–pituitary–thyroid axis regulation, as well as by the small sample size. The evidence base for reference values of thyroid hormones in rabbits remains unclear, and species-specific differences in molecular structure and receptor interactions compared with humans may exist³²; these factors were not fully accounted for in the present study.

Considering that a prior study showed significant effects of sex and age on thyroid 99mTcO₄ uptake in mice, which reflects NIS-mediated iodine trapping, sex and age were fixed in the present study³³. To our knowledge, comparable data in rabbits are currently unavailable. Although sex-related differences in tracer uptake may also exist in rabbits, our study demonstrated significant differences in iodine accumulation between experimental groups under conditions where sex was consistently controlled. Therefore, we believe that fixing sex as a controlled variable (either male or female) does not compromise the validity of intergroup comparisons or the overall conclusions of the study.

In the histopathological analysis, the high-dose groups identified as having iodine accumulation on DECT more frequently exhibited nodular hyperplasia, follicular atrophy, and inflammation, suggesting histopathological alterations; however, no statistically significant differences were observed between groups in the mean abnormality scores. These findings should be interpreted cautiously, because although a trend was observed in the high-dose group, the limited sample size may have reduced the statistical power to detect significant differences. In addition, discrepancies among iodine accumulation, hormonal changes, and histopathological findings may be attributable to a temporal lag between structural and functional alterations. In particular, histopathological changes may precede hormonal dysfunction during the subclinical phase of thyroid injury. Future studies should incorporate a longer follow-up period and a more finely graded dosing scheme to better elucidate the temporal relationships among iodine accumulation, histological damage, and endocrine function.

This study has several limitations. First, there is no standardized reference for thyroid hormone levels in rabbits, limiting the interpretation of thyroid function tests. The undetectable TSH levels across all groups also interfered evaluation of hypothalamic–pituitary–thyroid axis regulation. Second, although the administered high-dose ICM regimen (6.0 and 12.0 mL/kg, administered three times weekly) was intended to simulate cumulative high-dose exposure, the precise cumulative iodine dose could not be determined. Unlike previous studies in which cumulative dose could be directly quantified during procedures such as cardiac catheterization, the present study involved repeated ICM administration over an 8-week period. Consequently, continuous systemic elimination of iodine needed to be considered, which precluded accurate estimation of the total cumulative iodine exposure. Third, thyroid function was assessed only at baseline and at 4 and 8 weeks after ICM exposure, with no intermediate measurements, and rabbits were sacrificed immediately after the 8-week evaluation period. As a result, detailed short-term dynamics and longer-term thyroid functional changes could not be assessed.

In conclusion, repeated intravenous administration of high-dose iodinated contrast media in rabbits was associated with increased iodine accumulation in the thyroid gland, as quantified by DECT. However, the present study had limited ability to determine whether this iodine accumulation directly influences the incidence of subclinical hypothyroidism or induces histopathological changes.

Abbreviations

CT

Computed tomography

DECT

Dual-energy computed tomography

H&E

Hematoxylin and eosin

HU

Hounsfield unit

ICC

Intraclass correlation coefficient

ICM

Iodinated contrast media

ROI

Region of interest

VMI

Virtual monoenergetic imaging

TER

Thyroid enhancement ratio

TSH

Thyroid-stimulating hormone

Author contributions

S.L. and C.R. designed and led the study. C.R., S.L., and J.K. performed data curation and formal analysis. S.L. acquired funding and provided resources. C.R. drafted the original manuscript, and C.R., S.L., Y.H.C., J.Y.H., and J.E.C. reviewed and edited the final version.

Funding

This work was supported by Grant No. 0620215500 from Encormac (Seoul, Republic of Korea) and in part by the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (MSIT) (Grant No. RS-2023-00250062).

Data availability

The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.

Declarations

Competing interests

Seunghyun Lee is a Scientific Reports editorial board member but has no role in the review process or decision to publish this article. All remaining authors have declared no conflicts of interest.