Heliyon 9 (2023) e19509

Contents lists available at ScienceDirect

Heliyon
journal homepage: www.cell.com/heliyon

Radiation dose reduction and image quality evaluation for lateral

lumbar spine projection
Diego Nocetti a, *, Kathia Villalobos a, Nelson Marín b, Martina Monardes b,
Benjamín Tapia b, María Ignacia Toledo b, Camila Villegas b
a
Departamento de Tecnología Médica, Facultad de Ciencias de la Salud, Universidad de Tarapacá, Avenida 18 de septiembre N◦ 2222, 1010069,
Arica, Chile
b
Carrera de Tecnología Médica en Imagenología y Física Médica, Facultad de Ciencias de la Salud, Universidad de Tarapacá, Avenida 18 de
septiembre N◦ 2222, 1010069, Arica, Chile

A R T I C L E I N F O A B S T R A C T

Keywords: Purpose: Optimization studies in digital radiology help to reduce the radiological risk to patients
Digital radiology and maximize the benefits associated with their clinical purpose. The aim of this study was to
Dosimetry assess the optimization of lateral lumbar spine projection via a combination of exposure pa­
Image quality
rameters adjustments and additional filtration using a sectional anthropomorphic phantom.
Optimization
Materials and methods: We evaluated the effects of peak voltage, tube loading, and low-cost filters
made of copper, titanium, brass, and nickel on both the perceived and physical quality of 125
radiographs obtained in a computer radiography system. Signal-to-noise ratio (SNR) and contrast-
to-noise ratio (CNR) with their Figure of Merit (FOM), based on the entrance surface air kerma
with backscatter (ESAK), was used to assess physical image quality.
Results: The standard image had a perceived image quality, SNR, FOMSNR, CNR, FOMCNR and
ESAK of 3.4, 22.3, 386.4, 23.6, 433.7 and 1.28 mGy, respectively. Copper (90.3% purity) and
titanium (95.0% purity) filters reduced ESAK by an average of 60% without compromising
diagnostic quality, while brass and nickel filters increased dose under the conditions of the study.
Conclusions: Our findings show that optimizing lumbar spine projection can reduce radiation dose
without compromising image quality. Low-cost copper and titanium filters can be valuable in
resource-limited settings. Further research can explore additional strategies for radiological
optimization.

1. Introduction

Lumbar spine radiography is a widely used two-dimensional imaging procedure [1] that aids in the diagnosis and treatment of
pathologies or conditions that affect this area of the spine by providing images of both bony structures and soft tissues [2]. The quality
of the medical image plays a significant role in directly affecting the diagnosis and treatment of the patient’s condition [3]. Therefore,
the utilization of high-quality lumbar X-ray imaging can prevent missed diagnoses and enhance the accuracy of characterizing a range
of pathologies such as fractures, degenerative disc disease, lumbar spondylolisthesis, inter-articular structure damage and crack,
vertebral facet joint lesions, spinal stenosis, and tumors [4].

* Corresponding author.
E-mail address: [email protected] (D. Nocetti).

https://doi.org/10.1016/j.heliyon.2023.e19509
Received 3 April 2023; Received in revised form 29 July 2023; Accepted 24 August 2023
Available online 27 August 2023
2405-8440/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
D. Nocetti et al. Heliyon 9 (2023) e19509

The utilization of X-rays in radiological procedures has the potential to induce biological effects in exposed tissues, and result in
high radiation doses in comparison to other radiological examinations, thereby presenting a potential risk to the patient [5]. To
mitigate this risk, the radiation protection system advocates for the application of the optimization principle in medical practices
involving ionizing radiation [6,7]. This principle entails the use of strategies that generate images with adequate clinical utility at the
lowest achievable dose [5,8]. Consequently, with respect to radiation protection, optimizing exposure parameters is imperative to
minimize the likelihood of biological effects and to maximize the benefits associated with the procedures [5,9].
Due to their wider dynamic range, optimization is more practical to accomplish in digital radiography systems compared to analog
systems [10]. Studies have evaluated different optimization strategies, including factors like patient positioning [11], but with a
greater focus on altering exposure factors [12] and use of filtration [13]. Exposure values are responsible for modifying the energy and
intensity of the X-ray beam, while filters are employed to absorb the lower-energy X-rays, thereby elevating the average beam energy,
and reducing the dose to the patient’s skin [2,14,15].
Despite the importance of comprehensively examining the influence of multiple parameters on outcomes, along with the use of
both objective and subjective methods to evaluate their effect on image quality, only limited studies have addressed this matter [2,16].
Consequently, the purpose of this paper was to assess the optimization of lateral lumbar spine projection via a combination of exposure
parameters adjustments and additional filtration using a sectional anthropomorphic phantom.

2. Materials and methods

The evaluation was conducted at the Image Laboratory of the Universidad de Tarapacá in Arica, Chile, using a Shimadzu RAD speed
MF single-phase X-ray machine (Shimadzu Corp., Kioto, Japan) with a high-frequency generator. The digitizer system used was an
iCRco 3600-LF (iCRco Inc., CA, USA), which had a scanning resolution of 2.5 lp/mm, a matrix size of 1512 × 1260 pixels (with a pixel
size of 0.20 mm) and a bit depth of 16 bits.

2.1. X-ray system quality control

To ensure the accuracy and reliability of the X-ray system, we performed a basic quality control test, which included checking the
accuracy of peak voltage and exposure time, radiation output, and half-value layer. Additionally, we assessed the status of the
computed radiography system by conducting uniformity and linearity tests [17,18]. This aids in validating the technical operational
condition of the system, guaranteeing reproducibility of results, and minimizing the potential for incurring unnecessarily high doses to
obtain diagnostic image quality [19].

2.2. Filter characterization

For our optimization strategy, we considered four low-cost filters: copper (Marzú Ltd., Santiago, Chile), titanium (Fannie Tool Parts
Store, Guangzhou, China), brass (Aoshike Tech Store, Guangdong, China), and nickel (Magnet Expert, Guangzhou, China). We
evaluated the microstructure and elemental composition of each filter using scanning electron microscopy (SEM).
The samples were analyzed using a SEM Zeiss model EVO LS-10 (Carl Zeiss Meditec AG, Jena, Germany) without any coating. We
mounted the samples on aluminum holders with double contact carbon adhesives and analyzed them under low vacuum and variable
pressure mode (100 Pa in the chamber and 2 × 10− 5 Torr in the column). We used a working distance of 8.5 mm, a voltage of 20 kVp
and a tilt from 0◦ to 90◦ . For surface topography, we used magnifications of 1000 ×, and we acquired images in integration mode with
a resolution of 3024 × 2304 pixels. To perform elemental analysis of the samples, we used the Oxford EDS Inca program in SmartMap
mode, using energy-dispersive X-ray spectroscopy (EDX).
To assess the interaction of each material with X-rays generated at tube voltages ranging from 50 to 120 kVp (with effective energy
from 31.28 to 75.08 keV), we determined the mass attenuation coefficient using a 10X6-6 Radcal Model AccuDose ionization chamber
(Radcal Corp., CA, USA), vernier, and electronic scale.

2.3. Radiographic examination simulation

We acquired a total of 125 radiographs using five different tube voltages, five tube loading values, and five different filtration
conditions. Exposures were not performed with automatic exposure control (AEC) because this feature was not available. For the
standard image, we considered the exposure values typically used in the X-ray room evaluated in this study (75 kVp, 100 mAs).

Table 1
Experimental conditions used to optimize of the lateral lumbar spine radiographic examination.
Condition Peak voltages (kVp) Tube charges (mAs)

No filter 65; 70; 75; 80; 85 63; 80; 100; 125; 140
Copper 65; 70; 75; 80; 85 63; 80; 100; 125; 140
Titanium 65; 70; 75; 80; 85 63; 80; 100; 125; 140
Brass 75; 80; 85; 90; 95 90; 125; 140; 220; 320
Nickel 75; 80; 85; 90; 95 125; 140; 220; 250; 320

2
D. Nocetti et al. Heliyon 9 (2023) e19509

The literature reports a range of recommended peak voltages for this examination. For example, Geijer et al. [20] suggest using
values between 70 and 77 kVp, while Moore et al. [21], Ramli & Moey [9] and Alukic & Mekis [7] recommend the use of peak voltages
between 80 and 100 kVp, based on the probability of photoelectric absorption and Compton scattering of X-rays with the patient’s
tissues.
We selected the filter materials based on the proximity of their atomic number to copper, a material commonly used in radiological
equipment. Table 1 provides details of the experimental conditions and exposure values used in our study.
An RS-113T anthropomorphic pelvis and lumbar spine phantom (Radiology Support Devices, CA, USA) was used for the radio­
graphic examination simulations. This phantom is composed of materials that mimic bone structures embedded in a material that
simulates soft tissue. The phantom represents an average male patient (height 175 cm, weight 74 kg). We positioned the phantom
according to the study protocol proposed by Frank et al. [22] considering a constant focus-to-image receptor distance of 100 cm, as
shown in Fig. 1A. We used an X-ray grid from Shimadzu Corp. With a density of 40 L/cm and a grid ratio of 10:1. The interspacer
material between the lead strips in the grid was made of aluminum. Our image receptor had dimensions of 20 × 30 cm. To optimize the
exam, we adjusted the tube voltage, tube loading, and additional filter material.

2.4. Evaluation of perceived image quality

The perceived image quality was evaluated by 5 trained observers according to the European Guide’s criteria [23]. Four anatomical
structures, as depicted in Fig. 1B, were employed for evaluation. To ensure consistency with previous research, a 5-point scale [24–26]
was used to rate image quality and facilitate differentiation among conditions. The evaluation was performed in a single-blinded
fashion, with each image labeled by codes. The images were viewed on a ViewSonicVA2212m-LED monitor (ViewSonic Inc., Brea,
CA, USA), featuring a resolution of 1920 × 1080 and a refresh rate of 60Hz. Image analysis was conducted using ImageJ software
version 1.52a (National Institutes of Health, Bethesda, MD, USA). Illumination conditions were maintained for all image evaluations.

2.5. Assessment of physical image quality

We used the ImageJ software to assess the physical image quality by measuring the mean values and standard deviation of the
pixels contained within four regions of interest (ROIs) positioned in specific areas of the phantom. To reduce bias, all ROIs had the
same area. We collected data based on the methodology used by Lai et al. [2] (see Fig. 1C).
We evaluated three physical image quality parameters that are widely used in publications related to practice optimization in
radiology [27–29]:

⁃ Signal-to-noise ratio (SNR) that relates the useful information of the structure of interest (signal) to the variation in brightness level
within a ROI (noise) [30]. We calculated SNR using Equation (1) [27].
μ 1 − μROI 2
SNR = √ROI
̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
2 2
(1)
(σ ROI 1 ) +(σ ROI 2 )
2

⁃ Contrast-to-noise ratio (CNR) that corresponds to the ratio of the differences between the mean pixel values of the signal and
background, divided by the noise [31]. We estimated CNR using Equation (2) [32].

Fig. 1. Radiograph of the lateral projection of the lumbar spine. (A) Experimental setup for obtaining radiographs of the lateral lumbar spine. (B)
Anatomical points considered in the perceived image quality (1, spinous process; 2, pedicles and transverse processes; 3, upper and lower vertebral
plates; and 4, intervertebral disks). (C) Regions of interest (ROI) used to assess the physical image quality using ImageJ software.

3
D. Nocetti et al. Heliyon 9 (2023) e19509

μROI 1 − μROI 2
CNR = (2)
σ ROI 2
For formulas (1) and (2), we gathered information on pixels contained in ROI 1 from the center of L4 vertebra, and for ROI 2, we
considered only soft tissue from the anterior region of L4. μ corresponds to the mean value and σ to the standard deviation of the gray
value of the pixels contained in each ROI.

⁃ Figure of Merit (FOM) of the image quality metric (IQM), is an indicator of dose efficiency because it considers in only one
parameter the physical image quality and the ESAK delivered to the patient. We calculated FOMIQM using Equation (3).

IQM 2
FOMIQM = (3)
ESAK
We estimated this parameter by calculating the square of the IQM, corresponding to the SNR [27] or CNR [16], and dividing this
value by the ESAK used to generate each image.

2.6. Dosimetry

We measured the ESAK for each condition by placing an ionization chamber (10X6-6 Radcal Model AccuDose) within the irra­
diation field, aligned with the central beam and in contact with the phantom surface.

2.7. Statistical analysis

We tabulated and statistically analyzed data using the GraphPad Prism software version 7 (GraphPad, CA, USA). To assess the
interobserver agreement, we used the intraclass correlation coefficient (ICC) [33]. We tested the significance of differences between

Fig. 2. Characterization of the filter materials used to optimize the lateral projection of the lumbar spine: copper (A, B, C), titanium (D, E, F), brass
(G, H, I) and nickel (J, K, L)). (A, D, G, J) Scanning electron micrograph of surface topography (1000 × magnification). (B, E, H, K) Energy-dispersive
X-ray spectroscopy pattern of the elemental composition of the filter. (C, F, I, L) Mass attenuation coefficient versus effective energy for each
material. R2, coefficient of determination.

4
D. Nocetti et al. Heliyon 9 (2023) e19509

ESAK values using the one-way ANOVA test (with a 95% confidence level) or the Mann-Whitney U test, depending on the normality of
the variables determined by the Shapiro-Wilk test. We assessed the correlations in the scatterplots for perceived image quality by
calculating the Pearson correlation coefficient (r) [34]. Values of p < 0.05 were considered significant.

3. Results and discussion

In this study, we aimed to evaluate the efficacy of radiation dose reduction techniques and their impact on image quality for lateral
lumbar spine projection. Our results indicate that technical parameter adjustments, combined with filtration using low-cost copper and
titanium filters, can significantly reduce radiation dose while maintaining acceptable levels of image quality.
Quality control is an essential component to prevent excessive radiation doses to patients and should be the primary consideration
when starting an optimization process [6]. All quality control tests performed were within the tolerances specified in the Spanish
protocol for radiodiagnostic quality control (Table S1) [17]. Although crucial, this aspect is only minimally addressed in optimization
literature [35]. This article advocates for the inclusion of this dimension in scientific reports to validate the technical operational
condition of the system, ensure reproducibility of results, and facilitate comparison of outcomes achieved [30].
Given the usage of low-cost filter materials sourced from non-specialist vendors, it was deemed appropriate by the research team to
offer a more in-depth characterization of their surface microstructure and elemental composition (Fig. 2). Copper (Fig. 2A) and brass
(Fig. 2G) demonstrated the best state of preservation, while titanium exhibited some traces that may have resulted from sample
handling or storage (Fig. 2D). On the other hand, nickel displayed surface cracks (Fig. 2J).
The purity level and elemental composition of the samples are important factors in understanding X-ray interactions with each
material. To determine the purity level, EDX analysis showed a purity range between 83.5 and 95.0% (Fig. 2B, 2E, 2H, and 2K),
ensuring the use of representative samples for each chemical species. Furthermore, Fig. 2C, 2F, 2I, and 2L display the quantification of
the attenuation capacity of each sample, expressed as the mass attenuation coefficient, as a function of energy. For the assessed energy
range, the mass attenuation coefficient curves demonstrated coefficients of determination of at least 0.97. Copper exhibited the highest
attenuation, while titanium had the lowest (Fig. 2C, 2F, 2I, and 2L).
The samples exhibited varying percentages of impurities and surface irregularities, which, combined with the hardening of the
beam resulting from the polyenergetic nature of X-rays generated in a radiographic machine, may have slightly diminished the ac­
curacy of data fitting to the anticipated decreasing exponential trend [36].
Assessing the perception of human observers is a common method utilized in published optimization literature to evaluate image
quality [2,21]. This method has the advantage of approximating the clinical conditions under which human observers typically
evaluate images. However, the effect of potential observer subjectivity or variability can be assessed by determining the interobserver
agreement using the ICC [33]. Our study obtained excellent interobserver agreement (ICC 0.87–0.96), likely due to the observers’
similar training and experience.
Fig. 3 presents a summary of observer-perceived image quality for different filter conditions as a function of the ESAK used. Strong

Fig. 3. Perceived image quality in terms of entrance surface air kerma (ESAK) for lateral lumbar spine projection (A) without additional filtration,
and with additional filters of (B) copper, (C) titanium, (D) brass and (E) nickel. r, Pearson correlation coefficient (95% confidence interval); ICC,
intraclass coefficient correlation between observers (95% confidence interval). (F) Mean value ± standard deviation of the perceived image quality
for each filter condition. * p < 0.05; ****p < 0.0001.

5
D. Nocetti et al. Heliyon 9 (2023) e19509

to very strong positive correlations [34] were observed between these variables, with correlation coefficients ranging from 0.78 to 0.89
(Fig. 3A–E). Notably, the perceived image quality was significantly lower with the addition of brass filters compared to the series
without filters (Fig. 3F), consistent with findings by Lai et al. This result can be attributed to the increased photon fluence associated
with higher exposure values, which reduces noise perception and enhances image quality as a function of dose [2].
Fig. 4 illustrates the relationship between physical image quality and ESAK. The curves for SNR and CNR show a moderate increase
with increasing ESAK, while the curve for FOM exhibits a slight decrease, for both the filtered and unfiltered series (Fig. 4A–E).
As shown in Fig. 4G and I, this parameter showed significantly higher values in the series with added copper and titanium filtration
both for FOMSNR and FOMCNR, due to the lower patient dose [37]. The high attenuation owing to the thickness of the brass and nickel
foils led to increased technical factors to maintain the diagnostic value of the images [38]. However, the increase in dose did not
compensate for the change in image quality, with significantly lower FOMs obtained for both metrics in these materials.
Fig. 4F presents a comparative analysis of SNR, revealing significant differences (p < 0.0001) between the filtered and unfiltered
conditions. Copper and titanium filters showed significantly higher FOMSNR values than the unfiltered series, while brass and nickel
filters showed values approximately one order of magnitude lower than those of the unfiltered series (Fig. 4G). CNR and FOMCNR
values in Fig. 4H and I demonstrated similar behavior to that of SNR and FOMSNR, except for CNR with the titanium filter, which
showed values like those of the unfiltered series.
Since the brass and nickel filters excessively attenuated the X-ray beam, they produced inadequate image quality during the
preliminary tests. Therefore, in the experimental design, it was decided to increase the technical factors for these image series [34],
compared to the standard parameters (Table 1). However, the increase in dose did not compensate for the change in image quality,
resulting in significantly lower FOMs for both metrics in these materials. In contrast, copper and titanium filtration led to significantly
higher FOMSNR (Fig. 4G) and FOMCNR (Fig. 4I) values due to the lower dose for the patient [37].
Table 2 summarizes the best images obtained through optimization in perceived and physical image quality and ESAK reduction.
Exposure value modification yielded similar quality images to the standard with an ESAK reduction of about 24.7%. Copper and ti­
tanium filtration led to an additional ESAK reduction of up to 62.8%, resulting in up to 60.8% higher FOMSNR and up to 80.0% higher
FOMCNR than the standard values for images 1–6.
Regarding the use of filter materials, Table 2 shows that brass and nickel filters were ineffective in optimizing the procedure, as they
resulted in ESAK values 2.25 times higher than the standard image, with SNR and CNR values 50% lower and both FOMSNR and

Fig. 4. Physical image quality in terms of entrance surface air kerma (ESAK) for the lateral lumbar spine (A) without additional filtration and with
additional filters of (B) copper, (C) titanium, (D) brass and (E) nickel. (F) Signal-to-noise ratio (SNR) for each condition, (G) Figure of Merit of SNR
(FOMSNR), (H) Contrast-to-noise ratio (CNR) and (I) Figure of Merit of CNR (FOMCNR). Data are expressed as mean ± standard deviation. ****p
< 0.0001.

6
D. Nocetti et al. Heliyon 9 (2023) e19509

Table 2
Technical parameters (peak voltage, tube charge and filter condition), mean perceived image quality (IQ), physical image quality in terms of signal-
to-noise ratio (SNR), contrast-to-noise ratio (CNR) and their respective Figure of Merit (FOM) for the standard technique and the two best images
obtained with each condition under study.
Image Peak voltage (kVp) Tube charge (mAs) Filter condition Mean IQ ± SD SNR FOMSNR CNR FOMCNR ESAK variation (%)a

Standard 75 100 – 3.4 ± 0.5 22.27 386.36 23.60 433.72 –

1 85 63 – 3.4 ± 1.1 19.40 392.64 18.98 375.86 - 25.35
2 75 80 – 3.0 ± 0.0 21.56 476.25 21.41 469.38 - 23.96
3 75 100 Copper 3.0 ± 0.0 15.51 504.36 16.58 576.25 - 62.83
4 80 100 Copper 3.2 ± 0.8 17.38 510.57 17.46 515.66 - 53.93
5 75 80 Titanium 3.2 ± 0.4 17.45 621.15 18.78 719.82 − 61.84
6 70 100 Titanium 3.6 ± 0.5 17.30 586.77 19.95 780.49 − 60.27
7 90 140 Brass 3.0 ± 0.4 9.51 31.53 9.44 31.04 123.52
8 95 125 Brass 3.0 ± 0.5 9.84 29.99 10.17 32.01 151.64
9 85 140 Nickel 2.8 ± 0.8 9.48 38.95 9.89 42.33 79.83
10 75 320 Nickel 3.4 ± 0.5 12.06 55.00 13.34 67.30 106.07
a
Relative change respect to standard ESAK.SD, standard deviation; ESAK, entrance surface air kerma.

FOMCNR values about one tenth of those recorded in the standard image. In contrast, exposure value modification and the addition of
copper and titanium filters were successful in achieving significant reductions in ESAK, with values of 62.8% and 61.8% lower than the
standard condition, respectively. These findings highlight the importance of careful selection of filter materials in optimization pro­
cesses to achieve successful outcomes.
The X-ray attenuation characteristics of copper have been utilized in other optimization studies [2,14,39], and our findings align
with Martin’s results, where a 0.2 mm copper filter reduced the dose by more than 50% of the ESAK for abdominal radiography,
consisting of similar tissue to our study [38]. While no previous studies were found for titanium in this context, our promising results
suggest its potential use for optimization, supported by Fig. 4A–C which exhibit a clear separation between the curves, indicating a
broad optimization margin for these conditions. Conversely, brass and nickel filters failed to reduce the dose cost as explained above.
Fig. 4D and E showed a rapid overlap of SNR and CNR curves with their FOM curves, resulting in an unnecessarily high dose cost
compared to the standard condition.
The SNR trend shown in Fig. 4A–E is consistent with the direct proportionality between SNR and the square root of the dose used to
form the image [27]. Furthermore, the higher photon fluence of the unfiltered beam in comparison with that of the filtered series
(Fig. 4F) accounts for the significantly higher SNRs obtained for the former [38]. A similar trend is observed in CNR values, with the
unfiltered series showing higher values than the filtered series (Fig. 4H), owing to the reduction in photon fluence and beam hardening
resulting from the interposition of a filter material [36,38]. Moreover, as image quality is directly related to dose [2], the physical
image quality values obtained are specific to the imaging system characteristics and the technical parameters of each irradiation,
warranting caution when comparing studies [40].
The study by Lai et al. [2] achieved a dose reduction between 81% and 89% using a combination of optimization strategies,
including changes in peak voltage and tube loading, and the use of an added copper filter, in a digital radiography (DR) system.
However, their study differs from our work in terms of the technology used and the optimization reported. Also, they reported the
optimization as a function of the lower dose achieved, without considering the FOM of the estimated physical image quality, such as
the contrast-to-noise ratio (CNR), or the magnitude of the loss of perceived and physical image quality.
In a similar vein, Ramli and Moey [9] achieved dose reductions of almost 50% in a DR system by modifying only the peak voltage
and tube load values. However, it is worth noting that the reported ESAK values (26.63–54.95 mGy) are considerably higher than the
results obtained in our work, which ranged from 0.17 to 8.73 mGy. It is important to bear in mind that physical image quality values
depend on the imaging system’s characteristics and the technical parameters of each irradiation, thus, careful consideration should be
taken when comparing scientific articles.
The results obtained in this study provide a useful input for radiographers and medical personnel to optimize the lateral lumbar
spine radiographic examination, based on both objective and subjective evaluation of image quality. This is especially relevant for
medical facilities with limited resources, where low-cost filters such as copper and titanium can be employed to achieve significant
dose reduction without compromising image quality. Therefore, the findings of this study can have important implications for
improving patient care and safety, as well as for reducing healthcare costs associated with medical imaging.
It is crucial to acknowledge the limitations of the study. Firstly, the use of an anthropomorphic phantom may not adequately
capture the diverse variations present in real patients. Moreover, the utilization of a CR system, while suitable for the study, may have
inherent limitations when compared to alternative imaging systems. Furthermore, given the objective of employing cost-effective
materials, the purity of the samples becomes a significant factor that could potentially affect the generalizability of the results.

4. Conclusions

In this study, we demonstrated the efficacy of dose optimization through technical parameter adjustments and filtration in ESAK,
resulting in a significant reduction of 62.8% compared to the standard condition, without any significant reduction in physical or
perceptual image quality. These findings suggest that low-cost filters made of copper and titanium can be valuable tools for optimizing
radiation dose in medical facilities with limited resources for high-purity materials.

7
D. Nocetti et al. Heliyon 9 (2023) e19509

Moving forward, this study provides a foundation for further research exploring additional strategies for handling and interpreting
a greater number of simultaneous variables. Such strategies can be adapted and applied to other radiological systems, including flat
panel technologies. Future studies should aim to improve and refine the proposed methodology to ensure maximum patient benefit,
considering the clinical context, medical device characteristics, and anthropometric differences between subjects.

Author contribution statement

D. Nocetti: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed
reagents, materials, analysis tools or data; Wrote the paper.
K. Villalobos: Conceived and designed the experiments; Analyzed and interpreted the data; Contributed reagents, materials,
analysis tools or data; Wrote the paper.
N. Marín, M. Monardes, B. Tapia, M. Toledo, C. Villegas: Conceived and designed the experiments; Performed the experiments;
Analyzed and interpreted the data.

Data availability statement

The authors do not have permission to share data.

Funding statement

This research was partially supported by the Research Directorate at the Universidad de Tarapacá, through Senior Research Project
No. 7726-22.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Acknowledgments

The authors would like to thank the Bioarcheology Laboratory of the Universidad de Tarapacá and Professor Hipólito Núñez for
their assistance in taking the micrographs. We also thank to the Department of Medical Technology at the Universidad de Tarapacá
(Arica, Chile) for provide the access to the materials used in this research.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.heliyon.2023.e19509.

References

[1] A. Pazanin, D. Skrk, J.C. O’Driscoll, M.F. McEntee, N. Mekis, Optimal collimation significantly improves lumbar spine radiography, Radiat. Protect. Dosim. 189
(2020) 420–427, https://doi.org/10.1093/RPD/NCAA057.
[2] Z. Lai, C. Sá dos Reis, Z. Sun, Effective dose and image optimisation of lateral lumbar spine radiography: a phantom study, Eur. Radiol. Exp. 4 (2020) 1–9,
https://doi.org/10.1186/S41747-019-0132-3/TABLES/3.
[3] E. Kjelle, C. Chilanga, The assessment of image quality and diagnostic value in X-ray images: a survey on radiographers’ reasons for rejecting images, Insights
Imaging 13 (2022), https://doi.org/10.1186/S13244-022-01169-9.
[4] X. Chen, Q. Deng, Q. Wang, X. Liu, L. Chen, J. Liu, S. Li, M. Wang, G. Cao, Image quality control in lumbar spine radiography using enhanced U-net neural
networks, Front. Public Health 10 (2022) 1046, https://doi.org/10.3389/FPUBH.2022.891766/BIBTEX.
[5] N. Peacock, A. Steward, P. Riley, An evaluation of the effect of tube potential on clinical image quality using direct digital detectors for pelvis and lumbar spine
radiographs, J. Med. Radiat. Sci. 67 (2020) 260–268, https://doi.org/10.1002/JMRS.403.
[6] V. Tsapaki, Radiation dose optimization in diagnostic and interventional radiology: current issues and future perspectives, Phys. Med. 79 (2020) 16–21, https://
doi.org/10.1016/J.EJMP.2020.09.015.
[7] E. Alukic, N. Mekis, How does a non-optimal tube potential influence radiation dose to the patient in lumbar spine radiography? Radiography 27 (2021)
1105–1109, https://doi.org/10.1016/J.RADI.2021.04.014.
[8] International Atomic Energy Agency. Radiation protection and safety in medical uses of ionizing radiation, Specific Safety Standards Series SSG-46, IAEA,
Vienna, 2018, pp. 7–11.
[9] H. Ramli, S. Moey, Effects of kilovoltage on image quality and entrance surface dose in lumbar spine digital radiography, Iran, J. Med. Phys. 16 (2019) 368–371,
https://doi.org/10.22038/IJMP.2018.34167.1431.
[10] M. Uffmann, C. Schaefer-Prokop, Digital radiography: the balance between image quality and required radiation dose, Eur. J. Radiol. 72 (2009) 202–208,
https://doi.org/10.1016/j.ejrad.2009.05.060.
[11] A. Chaparian, A. Kanani, M. Baghbanian, Reduction of radiation risks in patients undergoing some X-ray examinations by using optimal projections: a Monte
Carlo program-based mathematical calculation, J. Med. Phys. 39 (2014) 32–39, https://doi.org/10.4103/0971-6203.125500.
[12] E. Gyan, S. Inkoom, G. Amoako, Optimal exposure factors for lumbar spine AP in computed radiography examinations, Int. J. Radiat. Res. 19 (2021) 421–427,
https://doi.org/10.52547/IJRR.19.2.21.

8
D. Nocetti et al. Heliyon 9 (2023) e19509

[13] A. Davis, S. Hopkins, Optimisation of patient dose for the horizontal beam technique in lateral lumbar spine radiographic examinations, Br. J. Radiol. 86 (2013),
20130053, https://doi.org/10.1259/bjr.20130053.
[14] E. Ekpo, A. Hoban, M. McEntee, Optimisation of direct digital chest radiography using Cu filtration, Radiography 20 (2014) 346–350, https://doi.org/10.1016/
J.RADI.2014.07.001.
[15] V. Hinojos-Armendáriz, S. Mejía-Rosales, M. Franco-Cabrera, Optimisation of radiation dose and image quality in mobile neonatal chest radiography,
Radiography 24 (2018) 104–109, https://doi.org/10.1016/J.RADI.2017.09.004.
[16] M. Freitas, R. Pimentel, L. Braga, F. Salido, R. Neves, R. Medeiros, Patient dose optimization for computed radiography using physical and observer-based
measurements as image quality metrics, Radiat. Phys. Chem. 172 (2020), 108768, https://doi.org/10.1016/J.RADPHYSCHEM.2020.108768.
[17] Spanish Society of Medical Physics, Protocolo Español de Control de Calidad en Radiodiagnóstico, 2011. Madrid.
[18] International Atomic Energy Agency, Protocolos de Control de Calidad para Radiodiagnóstico en América Latina y el Caribe, 2021.
[19] J. Zoetelief, R. van Soldt, I. Suliman, J. Jansen, H. Bosmans, Quality control of equipment used in digital and interventional radiology, Radiat. Protect. Dosim.
117 (2005) 277–282, https://doi.org/10.1093/RPD/NCI739.
[20] H. Geijer, E. Norrman, J. Persliden, Optimizing the tube potential for lumbar spine radiography with a flat-panel digital detector, Br. J. Radiol. 82 (2014) 62–68,
https://doi.org/10.1259/BJR/56572915.
[21] C. Moore, T. Wood, S. Balcam, L. Needler, T. Guest, W. Ngu, L. Chong, J. Saunderson, A. Beavis, Optimisation of tube voltage range (kVp) for AP abdomen,
pelvis and spine imaging of average patients with a digital radiography (DR) imaging system using a computer simulator, Br. J. Radiol. 93 (2020), https://doi.
org/10.1259/BJR.20200565/ASSET/IMAGES/LARGE/BJR.20200565.G002.JPEG.
[22] E. Frank, B. Long, B. Smith, Merrill’s Atlas of Radiographic Positioning & Procedures (Volume I), eleventh ed., Elsevier, Barcelona, 2010.
[23] European Commission, European Guidelines on Quality Criteria for Diagnostic Radiographic Images, 1996. https://op.europa.eu/en/publication-detail/-/
publication/d59ccc60-97ed-4ce8-b396-3d2d42b284be. June 14, 2022.
[24] H. Breit, J. Vosshenrich, V. Hofmann, T. Rusche, B. Kovacs, M. Bach, S. Manneck, D. Harder, Image quality of lumbar spine imaging at 0.55T low-field MRI is
comparable to conventional 1.5T MRI – initial observations in healthy volunteers, Acad. Radiol. 0 (2023), https://doi.org/10.1016/j.acra.2023.01.037.
[25] K. Chae, J. Goo, S. Ahn, J. Yoo, S. Yoon, Application of deconvolution algorithm of point spread function in improving image quality: an observer preference
study on chest radiography, Korean J. Radiol. 19 (2018) 147–152, https://doi.org/10.3348/KJR.2018.19.1.147.
[26] H. Mraity, A. England, S. Cassidy, P. Eachus, A. Dominguez, P. Hogg, Development and validation of a visual grading scale for assessing image quality of AP
pelvis radiographic images, Br. J. Radiol. 89 (2016), https://doi.org/10.1259/BJR.20150430/ASSET/IMAGES/LARGE/BJR.20150430.G003.JPEG.
[27] M. Oliveira, G. Lopez, P. Geambastiani, C. Ubeda, Application of newly developed Fluoro-QC software for image quality evaluation in cardiac X-ray systems,
Radiography 24 (2018) e44–e47, https://doi.org/10.1016/J.RADI.2017.12.006.
[28] A. Goode, C. Snyder, A. Snyder, P. Collins, M. DeLorenzo, P. Lin, Signal and contrast to noise ratio evaluation of fluoroscopic loops for interventional fluoroscope
quality control, J. Appl. Clin. Med. Phys. 20 (2019) 172–180, https://doi.org/10.1002/ACM2.12734.
[29] D. Abbeyquaye, S. Inkoom, N. Hammond, J. Fletcher, B. Botwe, Patient dose assessment and optimisation of pelvic radiography with computed radiography
systems, Radiat. Protect. Dosim. 195 (2021) 41–49, https://doi.org/10.1093/RPD/NCAB111.
[30] Z. Sun, C. Lin, Y. Tyan, K. Ng, Optimization of chest radiographic imaging parameters: a comparison of image quality and entrance skin dose for digital chest
radiography systems, Clin. Imag. 36 (2012) 279–286, https://doi.org/10.1016/J.CLINIMAG.2011.09.006.
[31] M. Mosleh-Shirazi, M. Amiri, R. Ravanfar Haghighi, M. Mahdavi, F. Zarei, Independent evaluation and comparison of digital radiography image quality in nine
major imaging centers affiliated to Shiraz University of Medical Sciences, Int. J. Radiat. Res. 19 (2021) 269–279, https://doi.org/10.52547/IJRR.19.2.4.
[32] S. Takagi, T. Yaegashi, M. Ishikawa, Dose reduction and image quality improvement of chest radiography by using bone-suppression technique and low tube
voltage: a phantom study, Eur. Radiol. 301 (30) (2019) 571–580, https://doi.org/10.1007/S00330-019-06375-6, 2019.
[33] T. Koo, M. Li, A guideline of selecting and reporting intraclass correlation coefficients for reliability research, J Chiropr Med 15 (2016) 155–163, https://doi.
org/10.1016/j.jcm.2016.02.012.
[34] P. Schober, L. Schwarte, Correlation coefficients: appropriate use and interpretation, Anesth. Analg. 126 (2018) 1763–1768, https://doi.org/10.1213/
ANE.0000000000002864.
[35] D. Nocetti, C. Ubeda, S. Calcagno, J. Acevedo, D. Pardo, Comparison of image quality among three X-ray systems for chest radiography: first step in
optimisation, Radiat. Protect. Dosim. 165 (2015) 386–391, https://doi.org/10.1093/rpd/ncv081.
[36] D. Chang, F. Lasley, I. Das, M. Mendonca, J. Dynlacht, Characteristics of photon beams, Basic Radiother. Phys. Biol. (2021) 67–73, https://doi.org/10.1007/
978-3-030-61899-5_7.
[37] A. Pavan, A. Alves, S. Duarte, G. Giacomini, T. Sardenberg, J. Miranda, D. Pina, Quality and dose optimization in hand computed radiography, Phys. Med. 31
(2015) 1065–1069, https://doi.org/10.1016/J.EJMP.2015.06.010.
[38] C. Martin, The importance of radiation quality for optimisation in radiology, Biomed. Imaging Interv. J. 3 (2006) e38, https://doi.org/10.2349/biij.3..e38.
[39] C. Charnley, A. England, A. Martin, S. Taylor, N. Benson, L. Jones, An option for optimising the radiographic technique for horizontal beam lateral (HBL) hip
radiography when using digital X-ray equipment, Radiography 22 (2016) e137–e142, https://doi.org/10.1016/J.RADI.2016.01.004.
[40] A. Brindhaban, K. Al Khalifah, G. Al Wathiqi, H. Al Ostath, Effect of x-ray tube potential on image quality and patient dose for lumbar spine computed
radiography examinations, Australas. Phys. Eng. Sci. Med. 28 (2005) 216–222, https://doi.org/10.1007/BF03178721.