Equal image quality and reduced radiation exposure in whole-spine X-ray imaging with slot-scanning technique compared with stitched radiography

Abstract

Purpose

Repeated whole-spine imaging in children and adolescents with scoliosis causes significant cumulative radiation exposure. Low-dose slot-scanning imaging can reduce radiation exposure. The aim of this study was to compare whole-spine imaging using the slot-scanning technique (SST) with conventional stitched radiography (SR) with regard to image quality and radiation dose.

Materials and Methods

We recruited 30 patients with a clinical indication for whole-spine imaging by SST, who had prior whole-spine imaging by SR available. 41 images acquired using both techniques were analyzed. The dose area product (DAP) and patient age, weight, height, and body-mass index were recorded. Image quality was assessed for 28 anatomical landmarks on a 4-point Likert scale by 4 raters. Group-wise comparisons were performed using paired t-tests. Interrater rating reliability was assessed using the interrater correlation coefficient.

Results

The mean DAP was significantly lower for SST than for SR (SST 8.0 ± 5.7 cGy · cm², SR 32.5 ± 34.5 cGy · cm², p < 0.001). The image quality was rated good for both modalities (means: SST 2.8 ± 0.6, SR 2.7 ± 0.5, p = 0.38). There was no significant difference in image quality ratings between both modalities. The inter-rater reliability was excellent with an inter-rater correlation coefficient of 0.86. Orthopedic surgeons rated the image quality of SST significantly better than that of SR.

Conclusion

This study shows a significant reduction in radiation exposure for slot-scanning whole-spine imaging compared to stitched radiography with equal image quality, which facilitates the reduction of radiation exposure for repeated whole-spine imaging in children and adolescents to 25%.

Key Points

• Image quality is comparable between slot-scanning and stitched whole-spine imaging.

• The slot-scanning technique had a significantly lower dose area product compared with stitched radiographs.

• Orthopedic surgeons rated the image quality of slot-scanning significantly better than that of stitching.

• Radiation exposure in scoliosis whole-spine imaging can be reduced to 25%.

Citation Format

• Liebscher H, Scherzer M, Meißner C et al. Equal image quality and reduced radiation exposure in whole-spine X-ray imaging with slot-scanning technique compared with stitched radiography. Rofo 2025; DOI 10.1055/a-2564-0580

 

MissingLinkText

Zusammenfassung

Ziel

Wiederholte Ganzwirbelsäulenaufnahmen bei Kindern und Jugendlichen mit einer Skoliose verursachen eine signifikante kumulative Strahlenbelastung. Slot-Scanning-Aufnahmen in Niedrigdosis können die Strahlenbelastung reduzieren. Das Ziel dieser Studie war der Vergleich von Ganzwirbelsäulenaufnahmen mittels Slot-Scanning-Technik (SST) und zusammengesetzten Röntgenaufnahmen (Stitching, SR) im Hinblick auf die Bildqualität und Strahlenbelastung.

Material und Methoden

30 Patienten mit einer klinischen Indikation für eine Ganzwirbelsäulenaufnahme mittels SST wurden rekrutiert, für die Voraufnahmen mittels SR vorlagen. Je 41 Aufnahmen in beiden Techniken wurden analysiert. Dosis-Flächen-Produkt (DFP), Alter, Gewicht, Größe und Body-Mass-Index wurden erfasst. Die Bildqualität wurde für 28 Landmarken auf einer 4-Punkt-Likert-Skala von 4 Ratern bewertet. Gruppenweise Vergleiche erfolgten mittels gepaartem t-Test. Die Inter-Rater-Reliabilität wurde mittels des Interrater-Korrelationskoeffizienten bewertet.

Ergebnisse

Das mittlere DFP war signifikant niedriger für SST als für SR (SST 8,0 ± 5,7 cGy · cm², SR 32,5 ± 34,5 cGy · cm², p < 0,001). Die Bildqualität wurde für beide Modalitäten gut bewertet (Mittelwerte: SST 2,8 ± 0,6; SR 2,7 ± 0,5, p = 0,38). Zwischen den Modalitäten gab es keinen signifikanten Unterschied in der Bildqualität. Die Inter-Rater-Reliabilität war exzellent mit einem Interrater-Korrelationskoeffizienten von 0,86. Orthopäden bewerteten die Qualität der SST signifikant besser als SR.

Schlussfolgerungen

Diese Studie zeigt eine signifikant niedrigere Strahlenbelastung bei Slot-Scanning-Ganzwirbelsäulenaufnahmen im Vergleich mit zusammengesetzten Aufnahmen bei gleichwertiger Bildqualität, was eine Reduktion der Strahlenbelastung für Kinder und Jugendliche bei wiederholten Ganzwirbelsäulenaufnahmen auf 25% ermöglicht.

Kernaussagen

• Die Bildqualität für Slot-Scanning- und zusammengesetzte Röntgenaufnahmen der ganzen Wirbelsäule ist vergleichbar.

• Slot-Scanning-Aufnahmen wiesen ein signifikant niedrigeres Dosisflächenprodukt als zusammengesetzte Röntgenaufnahmen auf.

• Orthopäden bewerteten die Qualität der Slot-Scanning-Aufnahmen signifikant besser als zusammengesetzte Aufnahmen.

• Die Strahlenbelastung durch Ganzwirbelsäulenaufnahmen bei Skoliosepatienten kann auf 25% reduziert werden.

 

MissingLinkText

Keywords

slot-scanning technique - scoliosis - whole-spine imaging - ALARA - pediatric radiographs - dose area product

Introduction

Scoliosis is a common condition in children and adolescents [1]. Whole-spine radiographic imaging is needed for the initial assessment of the degree of deformation and for follow-up during therapy. Whole-spine imaging results in radiosensitive organs like the mammary and thyroid glands being exposed to radiation. Radiation exposure in children and adolescents with scoliosis has been linked to an increased relative cancer risk with an odds ratio of 1.46 for cancer incidence and an odds ratio of 1.52 for cancer mortality compared to controls [2] [3] [4]. Therefore, steps to reduce radiation exposure must be taken, while taking the ALARA principle (as low as reasonably achievable) into account.

Conventional whole-spine imaging is achieved by stitching multiple separate radiographs of the spine together. There are two main problems with this method. Firstly, the cone shaped X-ray beam leads to a magnified and skewed projection of structures far from the center beam. Secondly, misalignment of stitching landmarks can lead to omission or duplication of vertebrae and can influence measurements [5].

Slot-scanning techniques use a nearly parallel collimated X-ray beam, which sequentially scans the whole spine, eliminating magnifying and skewing effects and the need for secondary stitching. A biplanar X-ray imaging system with simultaneous acquisition of anterior-posterior and lateral projections has already been reported to lead to markedly lower radiation exposure with comparable image quality to that of conventional radiography [6] [7] [8] [9].

Several anatomical landmarks are relevant for assessing the spine curvature and for Cobb angle measurements. Orthopedic spine surgeons and radiologists both assess whole-spine imaging, although in differing clinical contexts and with different aims.

The aim of our study was to compare the image quality and radiation exposure for whole-spine imaging using a monoplanar slot-scanning technique (SST) and stitched radiography (SR). We hypothesized that SST has at least an equal image quality rating compared with SR with reduced radiation exposure. Furthermore, we assessed differences in image quality ratings between orthopedic spine surgeons and radiologists.

 

MissingLinkText

Materials and Methods

The research protocol was approved by the local ethics committee. All research was conducted in compliance with the Declaration of Helsinki. Written informed consent was obtained from both patients, if feasible, and their legal guardian.

 

MissingLinkText

Patient recruitment

Patients of the outpatient scoliosis clinic of our hospital were screened for study participation. Patients were recruited if they had a clinical indication for follow-up whole-spine imaging, prior whole-spine imaging with stitched radiography was available, and there were no therapeutic interventions, i.e., spine surgery, between the imaging timepoints. After providing informed consent, the patients underwent follow-up imaging with the slot-scanning system. The matching prior imaging by stitched radiography was retrieved from the institute’s Picture Archiving and Communication System (PACS). All images were pseudonymized for rating. Patient age, sex, height, and weight, the calculated body-mass index (BMI), and the dose area product (DAP) as provided by the imaging system were recorded.

 

MissingLinkText

X-ray systems

Radiograph acquisition was done using a twin-robotic X-ray scanner with True-2-Scale whole-spine imaging and a standard digital radiography system with the stitching technique (Siemens Multitom Rax and Siemens Fluorospot Compact FD Ysio Max, respectively, Siemens Healthineers, Erlangen, Germany). A comparison of the technical details and imaging parameters of both imaging systems can be found in [Table 1]. Both systems are used at our institution in the daily clinical routine. The imaging parameters are in accordance with standard clinical practice. All radiographs were acquired in an upright standing position. An automatic exposure control is used for both systems. For comparison of both imaging systems and imaging parameters, we performed an X-ray phantom study, see Sup. 1, Sup. Tab. S1–3.

Table 1 Technical details for both imaging systems.

		Siemens Multitom Rax (T2S)	Siemens Fluorospot Compact FD Ysio Max
ap: anterior-posterior, lat: lateral
Input field size		42.0 × 42.5 cm	42.3 × 42.5 cm
Active detector matrix		2840 × 2874 px	2840 × 2874 px
Pixel size		296 µm (2 × 2 binning)	148 μm
Detector detective quantum efficiency [%], 2 μGy (RQA5)		70 % at 0.05 lp/mm 51 % at 1.0 lp/mm 42 % at 2.0 lp/mm 29 % at 3.0 lp/mm 19 % at Nyquist	67 % at 0.05 lp/mm 51 % at 1.0 lp/mm 42 % at 2.0 lp/mm 27 % at 3.0 lp/mm 18 % at Nyquist
Tube voltage	ap	< 10 y: 73 kV ≥ 10 y: 73 kV	< 8 y: 70 kV–77 kV ≥ 8 y: 73 kV–77 kV
	lat	< 10 y: 73 kV ≥ 10 y: 90 kV	< 8 y: 73 kV–77 kV ≥ 8 y: 73 kV–77 kV
Filter		0.3 mm Cu	0.2 mm Cu

 

MissingLinkText

Image Quality Evaluation

All pseudonymized radiographs were archived for later evaluation. Four raters – two radiologists (5 and 13 years of experience) and two orthopedic surgeons (14 and 13 years of experience) – were recruited and trained in rating imaging on a Likert scale from 1 to 4 on a set of training images (see Sup. Tab. S4 and Sup. Fig. S1). Following this training, the raters assessed anatomical landmarks adapted from current guidelines about quality criteria for diagnostic radiographs [10] [11]. 28 anatomic landmarks were evaluated in anterior-posterior images and 21 anatomic landmarks in lateral images (see Sup. Tab. S5) for all 82 images. Raters were blinded with regards to the imaging modality and image pairs.

 

MissingLinkText

Statistical Analysis

The image quality rating was calculated as the mean of all rating items for every image of each modality and rater. We defined a threshold of 0.5 for clinical relevance of differences in image quality ratings. The intermodal difference in image quality ratings was calculated as the difference between the image quality rating for each corresponding SST image and SR image (SST – SR). Groupwise comparisons of the image quality ratings and of the DAP were made using (paired) t-tests. A significance threshold of p < 0.05 was applied for all t-tests.

Interrater rating reliability (IRR) was assessed using the interrater correlation coefficient (ICC) with a two-way model, absolute agreement type and average-measure units. The interpretation of the ICC values was in accordance with Cicchetti (1994) as follows: poor < 0.4, fair 0.4–0.59, good 0.6–0.74, and excellent 0.75–1.0 [12].

All statistical analyses were performed using R version 4.3.3 (2024 The R Foundation for Statistical Computing) [13] and RStudio (2024.04.2 Build 764, 2009–2024 Posit Software, PBC). A list of R packages used for data analysis can be found in Supplement 2.

 

MissingLinkText

Results

[Table 2] shows the demographics of all patients. In total 30 patients were recruited. 30 radiographs were acquired in an anterior-posterior view and 11 radiographs in a lateral view, depending on the clinical context. Exemplary radiographs for both imaging methods are depicted in [Fig. 1]. Altogether, 41 pairwise image analyses were undertaken.

Table 2 Patient characteristics at the time of acquisition of the slot-scanning image.

*: number (percentage), §: mean ± standard deviation (minimum–maximum)
Sex*	Male	6 (20%)
	Female	24 (80%)
Age [years]§		15.9 ± 4.2 (5–21)
Weight [kg]§		49.2 ± 17.5 (10.9–80.0)
Height [cm]§		153 ± 22 (89–173)
BMI [kg/m²]§		20.3 ± 6.2 (10.2–44.7)

Fig. 1 Exemplary radiographs using the stitching technique a, c and slot-scanning technique b, d. For the first patient a/b, there is a clear delineation of all vertebral contours in both techniques. For the second patient c/d, there is a better delineation of vertebral contours with the slot-scanning technique compared with the stitching technique. Dose area products [cGy 1 · cm²]: a) 22.6 b) 7.1 c) 8.9 d) 3.3.

Dose area product

The DAP was significantly lower for SST (mean 8.0 cGy · cm², standard deviation [SD] 5.7 cGy · cm²) than for the SR (mean 32.5 cGy · cm², SD 34.5 cGy · cm²) with a mean difference (SR – SST) of 24.4 cGy · cm² and a mean ratio (SR / SST) of 4.1 (t(40) = 4.8, p < 0.001), see [Fig. 2] and [Table 3]. The average DAP was higher for lateral projections than for anterior-posterior projections for both methods. The DAP increased exponentially with patient weight and BMI, see Supplemental Figure S2.

Fig. 2 Distribution of dose area products (DAP) for stitched radiographs (SR) and slot-scanning technique (SST).

Table 3 Summary of dose area products (cGy · cm²) by method and projection.

Method	Projection	Mean	SD	Min.-max.
SST: slot-scanning technique, SR: stitched radiograph, ap: anterior-posterior, lat: lateral, SD: standard deviation, min.: minimum, max.: maximum
SST	Overall	8.0	5.7	2.0–25.9
	ap	7.5	5.3	2.0–22.0
	lat	9.6	6.8	2.1–25.9
SR	Overall	32.5	34.5	1.0–134.0
	ap	28.5	29.5	1.0–127.7
	lat	43.4	45.3	1.7–134.0

 

MissingLinkText

Image quality

The image quality ratings were on average good for both SST and SR (SST: mean 2.8, SD 0.6; SR: mean 2.7, SD 0.5), see [Table 4] and [Fig. 3]. There was no significant difference in image quality ratings between both modalities (t(40) = –0.89, p = 0.38). Subgroup analysis by projection showed that image quality ratings were on average 0.5 score units higher for the anterior-posterior projections than for the lateral projections (a.-p.: mean 2.9; lat.: mean 2.4) with a significant between-group difference (t(30.9) = 4.31, p < 0.001), see [Table 4] and Sup. Fig. S3. There were no significant differences between SST and SR for both projections.

Table 4 Image quality ratings by method and projection.

Method	Projection	Mean	SD	Min.–Max.
SST: slot-scanning technique, SR: stitched radiograph, ap: anterior-posterior, lat: lateral, SD: standard deviation, min.: minimum, max.: maximum
SST	Overall	2.8	0.6	1.6–3.7
	ap	2.9	0.5	1.9–3.7
	lat	2.4	0.5	1.6–3.3
SR	Overall	2.7	0.5	1.5–3.5
	ap	2.9	0.3	1.9–3.5
	lat	2.3	0.6	1.5–3.3

Fig. 3 Comparison of image quality ratings for stitched radiographs (SR) and slot-scanning technique (SST). Dashed lines connect paired values between methods.

The intermodal difference in image quality ratings was on average very small (mean 0.04), see [Table 5] and [Fig. 4]. The intermodal difference was on average slightly higher for the lateral projections (mean 0.14) compared with the anterior-posterior projections (mean 0.006), see [Table 5] and [Fig. 4], without statistical significance (t(15.7) = –1.18, p = 0.25). For most images, the intermodal differences in image quality ratings were within the predefined threshold for clinical irrelevance (37/41 = 90%).

Table 5 Intermodal difference in image quality ratings (SST – SR) by projection.

Projection	Mean	SD	Min.–max.
ap: anterior-posterior, lat: lateral, SD: standard deviation, min.: minimum, max.: maximum
Overall	0.04	0.3	–0.6–0.6
ap	0.01	0.3	–0.6–0.6
lat	0.14	0.3	–0.5–0.5

Fig. 4 a Distribution of intermodal differences in image quality ratings for all patients. b Distribution of intermodal differences in image quality ratings by projection. ap: anterior-posterior, lat: lateral, SR: stitched radiographs, SST: slot-scanning technique.

There were differences in the image quality ratings between the several target vertebrae for both imaging methods, see Sup. Fig. S4. Image quality ratings were considerably lower for the C2 vertebra compared with all other vertebrae in the anterior-posterior projection for both imaging modalities (means: C2 1.5, C7 3.2, Th7 2.9, Th12 3.0, L1 3.0, L5 2.9). In contrast, image quality ratings were higher on average in the lateral projection for the C2 vertebra as well as for the L5 vertebra compared with the other target vertebrae in both imaging methods (means: C2 3.0, C7 2.4, Th7 1.8, Th12 2.1, L1 2.3, L5 2.9). Image quality ratings were generally similar between slot-scanning radiographs and stitched radiographs for most items. There were two notable exceptions: The ratings for the C2 vertebra in the lateral projection were substantially higher for slot-scanning technique radiographs than for stitched radiographs (SST 3.4, SR 2.6). Also, the ratings for the femoral heads were substantially higher for slot-scanning radiographs than for stitched radiographs in both the anterior-posterior and lateral projection (a.-p.: SST 3.3, SR 1.7; lat.: SST 2.2, SR 1.4).

There was only a slight influence of patient weight on the image quality ratings. Image quality ratings were lower for both extremely low and high BMI values, see Sup. Fig. S5.

 

MissingLinkText

Inter-rater reliability

The inter-rater reliability with regard to the rating of individual items was high. As depicted in [Fig. 5], there was a good correlation between the average ratings and the individual ratings for each radiograph. The ICC was 0.86, indicating excellent agreement between raters in their ratings of individual items. One rater assigned on average slightly lower image quality than all other raters, see Sup. Fig. S6.

Fig. 5 a Agreement between raters: Distribution of average ratings of all raters and individual ratings for each rater. b Image quality ratings by rater specialty and imaging method. SR: stitched radiographs, SST: slot-scanning technique.

Raters from both specialties (radiology and orthopedics) assigned on average similar ratings for slot-scanning technique radiographs (means: radiology 2.8, orthopedics 2.8), but different ratings for the stitched radiographs (means: radiology 3.0, orthopedics 2.5). Consequently, there was a significant difference of the intermodal difference in image quality ratings between radiologists and orthopedic surgeons (means: radiology –0.2, orthopedics 0.3, t(81) = 10.8, p < 0.001).

 

MissingLinkText

 

MissingLinkText

Discussion

This study compared the DAP and image quality of whole-spine radiographs with SR and SST in an intra-individual approach. We found significantly lower DAPs in both anterior-posterior and lateral views when the slot-scanning technique was used while image quality was equal for both techniques. Ratings of slot-scanning radiographs did not differ between radiologists and orthopedic surgeons.

The DAP for SST was on average only 25% of the DAP for SR with no significant difference in image quality ratings. Although the DAP is only an approximation for radiation exposure, it is widely used in clinical practice. These findings align with the literature on a similar biplanar slot-scanning system analyzing both DAP and skin entrance dose. Here, Yvert et al. found superior image quality ratings for the slot-scanning system compared with standard digital radiography with no significant difference in DAP, but lower skin entrance doses for the slot scanning system for thoracic and lumbar regions with ratios of 1.49 and 2.15, respectively [9]. Deschênes et al. and Hui et al. found a significant reduction in radiation exposure for a slot-scanning system, with ratios of 2.9–9.2 and 5.6–26.7, respectively, and equal or increased image quality [6] [7]. In standard radiography, a dose reduction of 45% for a dose-reduced protocol compared with the normal-dose protocol did not lead to worse image quality [14].

Differences in dose between the two imaging modalities could be attributed to X-ray system-inherent factors. Generally, the detector’s physical properties were nearly identical. With our phantom measurements, we could exclude a system-inherent dose-reducing effect favoring the slot-scanning system. As shown in our phantom measurements, the use of a slightly thicker copper filter and a higher tube voltage led to a dose reduction. However, the magnitude of the dose reduction was in both cases much smaller than the observed total dose reduction. The main effect of the dose reduction can, therefore, be attributed to the slot-scanning technique. Consequently, by choosing the appropriate imaging technique and parameters with dose-adapted protocols, one can reduce radiation exposure without lowering image quality. An even higher dose reduction seems possible if a generally available microdose protocol is used. However, this microdose protocol was not used for the patients included in this study. This option needs to be investigated further.

X-ray image quality depends on two factors: image contrast and noise. Image contrast mainly depends on the energy of the X-ray photons, with photons of lower average energy providing more image contrast with reduced penetration. Generally, both imaging methods used similar tube voltages with similar image contrast. For lateral imaging, the slot-scanning system used a slightly higher tube voltage in older children, increasing penetration, but decreasing image contrast. Noise is generated by scattering events, which can be reduced, e.g., by grids and collimation. One downside of anti-scatter grids is the need for increased exposure due to partial X-ray absorption, which increases dose and is therefore not favored in pediatric imaging [15]. Compared with conventional stitched X-rays, the slot-scanning system uses an x-ray-beam with a high collimation along the patient’s vertical axis. This lowers the likeliness of scatter events due to the smaller irradiated area. Luckner et al. showed in a phantom study that the slot-scanning approach showed better signal-to-noise and scatter-to-primary ratios than the conventional full-field acquisition with estimated dose reduction potentials of 28% to 67% for equal image quality [16].

Generally, the mean image quality was good for both methods with no significant differences between methods, which enables adequate assessment of all relevant anatomic features. For the vast majority of images, small intermodal differences in image quality ratings are without clinical relevance. There was a non-significant difference in image quality ratings of the lateral projections only, slightly favoring the SST. This was mostly due to a better delineation of the C2 vertebra using SST compared with SR. This is presumably due to both a higher local radiation dose and differences in image projection, as the C2 vertebra is far off from the center beam of the stitched radiographs. Similar observations of equal or even higher image quality ratings have been made for a biplanar slot-scanning system compared with standard radiography. Yvert et al. reported a higher image quality for slot-scanning radiographs compared with standard digital radiographs for the cervical and thoracic spine and lower ratings for the lumbar spine [9]. Similarly, Deschênes et al. reported an equal or better global image quality for a slot-scanning system compared with a standard computed radiography system and generally higher image quality in the cervical spine [6]. Corresponding findings were obtained for a comparison of a standard dose protocol and reduced dose protocol for a standard digital radiography system with no significant difference for several image quality parameters [14]. In summary, our findings agree with the literature on comparable or better image quality in dose-reduced slot-scanning radiographs of the spine compared with standard radiography.

Regardless of the imaging method, image quality was not homogenous along the whole spine but was affected by patient-inherent factors like body weight, region-specific anatomy, and scoliosis characteristics. We found lower image quality ratings for the C2 vertebra compared with all other target vertebrae in anterior-posterior projection, which is mainly due to overlying structures of the viscerocranium. Conversely, we found higher image quality ratings for the C2 and L5 vertebrae in the lateral projection compared with the other vertebrae for both imaging methods, presumably due to fewer overlying structures in the cervical and lumbar spine and a higher degree of vertebral rotation due to scoliosis in the thoracic spine. Morel et al. similarly reported higher image quality ratings for the cervical and lumbar spine on lateral view imaging for both low-dose and micro-dose slot-scanning whole-spine imaging [8], representing a strength of SST. Incidentally, we found higher image quality ratings for the femoral heads for SST than for SR in both projections. This is presumed to be the result of differences in image collimation, as the SST had a wider lateral collimation.

Most comparisons of SST in the literature used matched-group analyses without a direct comparison of the imaging methods on an patient-individual level [7] [9] [14], with the exception of Deschênes et al., who consecutively acquired both SR and SST [6]. One strength of our study was the intra-individual comparison of both imaging methods, which facilitates statistical correction of inter-individual variance. SR and SST were not acquired simultaneously, but at different timepoints during clinical follow-up. Therefore, no additional study-related radiation exposure occurred. As the potential time-dependent intra-individual variability was assumed to generally be lower than the inter-individual variability, we presume adequate matching between groups. Additionally, it was ensured during patient recruitment that no relevant therapeutic interventions, especially spine surgery, took place between both imaging timepoints to ensure comparability between both imaging methods.

Image quality was mainly independent of weight and BMI with the exception of both a high and a low BMI. This appears to be due to the automated dose adaption for both modalities, ensuring adequate imaging quality over a large range of different patient weights.

Our study showed excellent agreement between the four different raters with an ICC of 0.86. Only one rater showed a slight deviation in average image quality ratings from the average over all raters with generally lower scores for both modalities. These findings match reports in the literature of high inter-rater reliability measures. Deschênes et al. showed generally good agreement for posterior-anterior projections with differences in the lateral projections for the lumbar region [6]. Hui et al. reported ICCs of 0.883–0.942 and Ernst et al. reported ICCs of 0.934–0.973, indicating excellent inter-rater agreement, whereas Yvert et al. reported ICCs of 0.35–0.58 [7] [9] [14]. Importantly, the orthopedic surgeons rated the image quality of slot-scanning radiographs significantly higher than stitched radiographs. These results imply that the slot-scanning technique adds substantial value in the clinical routine for orthopedic surgeons when planning and monitoring scoliosis surgery.

Our study has some limitations. There was a slight difference in image quality ratings between radiologists and orthopedic surgeons for SR, but not for the SST. The difference may be attributable to one rater, who generally assigned lower ratings than all other raters. Furthermore, the consecutive differences in intermodal differences were not clinically relevant. Additionally, although raters were formally blinded with regards to imaging modality during reviews, a partial unblinding due to modality-inherent differences in general image impression cannot be excluded. As there was no definite trend in intermodal differences in image quality ratings, a relevant bias of all raters towards one modality due to partial unblinding is unlikely.

 

MissingLinkText

Conclusion

This study compared whole-spine radiography for the imaging of scoliosis with a slot-scanning system and standard stitched radiography. We found an average reduction of radiation exposure to 25% using slot-scanning radiography without compromising image quality. Therefore, between the two examined imaging systems, the slot-scanning system can be considered a means to reduce radiation exposure for repeated imaging in children and adolescents in clinical practice in line with the ALARA principle.

 

MissingLinkText

Clinical relevance of the study:

• Whole-spine imaging by a slot-scanning system can facilitate a significant dose reduction with equal imaging quality compared with a stitched radiography system.

• Dose reduction is especially relevant in children and adolescents due to higher radiation absorption and larger proportions of radiosensitive tissue.

• Slot-scanning techniques offer additional benefits compared with stitched radiography with regards to projection geometry and image quality of certain critical structures.

 

MissingLinkText

 

MissingLinkText

Conflict of Interest

SFUB has received lecture fees from Siemens Healthineers. UP has received payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events from Stryker and support for attending meetings and/or travel from Stryker, Medtronic, Nuvasive, Globus Medical.

Acknowledgement

We wish to thank Ms T. Fritzsche for her contributions to data acquisition.

• References

• 1 Weinstein SL, Dolan LA, Cheng JC. et al. Adolescent idiopathic scoliosis. The Lancet 2008; 371: 1527-1537
• 2 Morin Doody M, Lonstein JE, Stovall M. et al. Breast Cancer Mortality After Diagnostic Radiography: Findings From the U.S. Scoliosis Cohort Study. Spine 2000; 25: 2052-2063
• 3 Law M, Ma WK, Lau D. et al. Cumulative radiation exposure and associated cancer risk estimates for scoliosis patients: Impact of repetitive full spine radiography. Eur J Radiol 2016; 85: 625-628
• 4 Luan FJ, Wan Y, Mak KC. et al. Cancer and mortality risks of patients with scoliosis from radiation exposure: a systematic review and meta-analysis. Eur Spine J 2020; 29: 3123-3134
• 5 Supakul N, Newbrough K, Cohen MD. et al. Diagnostic errors from digital stitching of scoliosis images – the importance of evaluating the source images prior to making a final diagnosis. Pediatr Radiol 2012; 42: 584-598
• 6 Deschênes S, Charron G, Beaudoin G. et al. Diagnostic Imaging of Spinal Deformities: Reducing Patients Radiation Dose With a New Slot-Scanning X-ray Imager. Spine 2010; 35: 989-994
• 7 Hui SCN, Pialasse JP, Wong JYH. et al. Radiation dose of digital radiography (DR) versus micro-dose x-ray (EOS) on patients with adolescent idiopathic scoliosis: 2016 SOSORT- IRSSD “John Sevastic Award” Winner in Imaging Research. Scoliosis Spinal Disord 2016; 11: 46
• 8 Morel B, Moueddeb S, Blondiaux E. et al. Dose, image quality and spine modeling assessment of biplanar EOS micro-dose radiographs for the follow-up of in-brace adolescent idiopathic scoliosis patients. Eur Spine J 2018; 27: 1082-1088
• 9 Yvert M, Diallo A, Bessou P. et al. Radiography of scoliosis: Comparative dose levels and image quality between a dynamic flat-panel detector and a slot-scanning device (EOS system). Diagn Interv Imaging 2015; 96: 1177-1188
• 10 European Commission, Directorate-General for Research and Innovation. European guidelines on quality criteria for diagnostic radiographic images in paediatrics. Publications Office. 1996
• 11 Bundesärztekammer, Hrsg. Leitlinie der Bundesärztekammer zur Qualitätssicherung in der Röntgendiagnostik. Dtsch Ärztebl 2023; 120.
• 12 Cicchetti DV. Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology. Psychol Assess 1994; 6: 284-290
• 13 R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2024
• 14 Ernst C, Buls N, Laumen A. et al. Lowered dose full-spine radiography in pediatric patients with idiopathic scoliosis. Eur Spine J 2018; 27: 1089-1095
• 15 Uffmann M, Schaefer-Prokop C. Digital radiography: The balance between image quality and required radiation dose. Eur J Radiol 2009; 72: 202-208
• 16 Luckner C, Weber T, Herbst M. et al. A phantom study on dose efficiency for orthopedic applications: Comparing slot-scanning radiography using ultra-small-angle tomosynthesis to conventional radiography. Med Phys 2020; mp.14680

Publication History

Received: 29 August 2024

Accepted after revision: 19 March 2025

Article published online:
25 April 2025