- Email: [email protected]
Application of newly developed Fluoro-QC software for image quality evaluation in cardiac X-ray systems
Radiography xxx (2017) 1e4
Contents lists available at ScienceDirect
Radiography journal homepage: www.elsevier.com/locate/radi
Application of newly developed Fluoro-QC software for image quality evaluation in cardiac X-ray systems M. Oliveira a, *, G. Lopez a, P. Geambastiani b, C. Ubeda c a
Department of Heath Technology and Biology, Federal Institute of Bahia, Emídio dos Santos e s/n, Salvador, Bahia, Brazil Cardio Pulmonar Hospital, Av. Anita Garibaldi, 2199 - Garibaldi, Salvador, Bahia, Brazil c Medical Technology Department, Health Sciences Faculty, Tarapaca University, Av. 18 de septiembre, n 2222, Arica, Chile b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 3 August 2017 Received in revised form 9 December 2017 Accepted 13 December 2017 Available online xxx
Introduction: A quality assurance (QA) program is a valuable tool for the continuous production of optimal quality images. The aim of this paper is to assess a newly developed automatic computer software for image quality (IR) evaluation in fluoroscopy X-ray systems. Methods: Test object images were acquired using one fluoroscopy system, Siemens Axiom Artis model (Siemens AG, Medical Solutions Erlangen, Germany). The software was developed as an ImageJ plugin. Two image quality parameters were assessed: high-contrast spatial resolution (HCSR) and signal-tonoise ratio (SNR). The time between manual and automatic image quality assessment procedures were compared. The paired t-test was used to assess the data. p Values of less than 0.05 were considered significant. Results: The Fluoro-QC software generated faster IQ evaluation results (mean ¼ 0.31 ± 0.08 min) than manual procedure (mean ¼ 4.68 ± 0.09 min). The mean difference between techniques was 4.36 min. Discrepancies were identified in the region of interest (ROI) areas drawn manually with evidence of user dependence. The new software presented the results of two tests (HCSR ¼ 3.06, SNR ¼ 5.17) and also collected information from the DICOM header. Significant differences were not identified between manual and automatic measures of SNR (p value ¼ 0.22) and HCRS (p value ¼ 0.46). Conclusion: The Fluoro-QC software is a feasible, fast and free to use method for evaluating imaging quality parameters on fluoroscopy systems. © 2017 The College of Radiographers. Published by Elsevier Ltd. All rights reserved.
Keywords: Image quality Fluoroscopy Image metrics Quality control
Introduction A quality assurance (QA) program is a valuable tool for the continuous production of optimal quality images. Uncalibrated equipment could produce insufficient image quality (IQ) and lead to an increase in radiation dose to patients and staff.1 In some cases, maintaining an acceptable level of IQ in interventional fluoroscopy procedures requires a substantial increase in radiation dose per frame.2 IQ is currently assessed for contrast, resolution and artefacts. For an X-ray system, the signal-to-noise ratio (SNR) is directly proportional to the square root of the X-ray dose used to create the image.3 SNR is the ratio of useful information by the standard deviation (SD) of the grey level (noise) value.4 Noise is derived from * Corresponding author. E-mail addresses: [email protected] (M. Oliveira), [email protected] (G. Lopez), [email protected] (P. Geambastiani), [email protected] (C. Ubeda).
the quantum noise properties of X-ray photons and the electronic noise (unrelated to the number of photons detected) of the detection system.5 Both noise sources affect quality control test results. There are no detailed studies investigating the IQ optimization of the X-ray equipment used in interventional radiology.6 The aim of this paper is to assess a newly developed automatic computer software based method for image quality evaluation in fluoroscopy systems. Materials and methods Description of Fluoro-QC macro The Fluoro-QC macro code was developed for ImageJ to analyse IQ for the fluoroscopy technique. ImageJ7 is open-source software and thus it does not require a user licence. The program works independently of the operating system. Fluoro-QC works as a postprocessing tool, although the original image is not manipulated (Fig. 1).
https://doi.org/10.1016/j.radi.2017.12.006 1078-8174/© 2017 The College of Radiographers. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Oliveira M, et al., Application of newly developed Fluoro-QC software for image quality evaluation in cardiac Xray systems, Radiography (2017), https://doi.org/10.1016/j.radi.2017.12.006
2
M. Oliveira et al. / Radiography xxx (2017) 1e4
Figure 3. Experimental setup of the PMMA phantom and test object.
The macro creates four rectangular regions of interest (ROI) on the DICOM image acquired in cine mode. The numerical IQ evaluation is always performed on three consecutive images for each series located at sequential positions 10, 12 and 15, so as to avoid the stability problems likely to occur in the first images of the series. The ROIs are positioned in the same place as in the study performed by Vano et al.8 IQ is evaluated by analysing the lowcontrast circles and the high-contrast spatial resolution (HCSR) groups. The parameters evaluated are: SNR, HCSR and one figure of merit (FOM) (Fig. 2). To evaluate SNR, the following calculation is used:
Figure 1. Software workflow.
½BG ROI1 ffi SNR ¼ rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðSD2ROI þSD2BG Þ
(1)
2
where ROI1 is the mean value of the pixel inside circle number 3 and BG is the background mean value of the pixel placed near the low-contrast circles. The standard deviation is SDROI for ROI1, while SDBG is the standard deviation for BG. The following equation is used to numerically evaluate HCSR:
HCSR ¼ SD3 SD4
(2)
where SD3 is the standard deviation for the pixel content in the ROI3, inside the eighth group in the central grid of the pattern bar, and SD4 is the standard deviation for the pixel content in the ROI4
Figure 2. An illustration of the four ROI's automatically applied to specific places on the fluoroscopy image from the TOR 18 FG test object.
Figure 4. Comparison of timing requirements for the manual and automatic (FluoroQC) QC process.
Please cite this article in press as: Oliveira M, et al., Application of newly developed Fluoro-QC software for image quality evaluation in cardiac Xray systems, Radiography (2017), https://doi.org/10.1016/j.radi.2017.12.006
M. Oliveira et al. / Radiography xxx (2017) 1e4
3
Table 1 HCRS, SNR and FOM results according to manual and automatic ROIs. Manual
Automatic
Area (pixels2)
Area (pixels2)
ROI 1 ROI 2 ROI 3 ROI 4 SNR User “A” 18.9 User “B” 21.4 User “C” 16 Mean and standard deviation
21.8 26.8 10
18.4 22.8 8.9
7.1 23.4 13.4
5.27 5.62 5.25 5.38 ± 0.17
HCSR FOM 3 3.5 3.1 3.2 ± 0.22
ROI 1 ROI 2 ROI 3 ROI 4 SNR HCSR FOM
0.012 15 0.013 15 0.011 15
15 15 15
placed in the periphery of the high-contrast groups and representative of the noise in this area. FOM was evaluated and defined as:
FOM ¼
SNR2 DAP
p-Value manual p-Value manual and automatic and automatic mode for HCRS mode for SNR
(3)
where DAP is the dose area product obtained by the DICOM header. The user did not need to manually choose the frame number because it was always selected (automatically) by the frame representing the middle of the acquisition series. Evaluation of Fluoro-QC Three users with considerable experience in IQ evaluation calculated SNR and HCSR on the same image, both manually and using Fluoro-QC. A timer was used in order to compare the time consumed between the two procedures. Time consumed was counted from image selection until the presentation of results. Images were obtained for one X-ray system, the Siemens Axiom Artis model (Siemens AG, Medical Solutions Erlangen, Germany). Analysis was performed on images in the standard reduced format of 512 512 and 8 bits. Fig. 3 shows the experimental arrangement with the polymethyl methacrylate (PMMA) phantom and physical test object (Leeds TOR 18-FG).9 Statistical analysis Microsoft Office Excel software was used for statistical analysis and the paired t-test was used to assess the data. p values of less than 0.05 were considered significant. Results The results of the IQ assessment and the information from the DICOM header (manufacturer, the number of frames, tube voltage (kV), tube current (mA), exposure time(s), dose area product, intensifier size, matrix size, radiation mode and source-to-patient distance) were presented in a final report. Fig. 4 shows the time comparison between the automatic and manual tests. Significant differences were not demonstrated between the manual and automatic QC methods for SNR (p ¼ 0.22) and HCRS (p ¼ 0.46) values. The results of HCRS and SNR in Manual mode were influenced by discrepant values of ROIs size (Table 1). Three tests were performed using Fluoro-QC and results were obtained more quickly (0.26e0.43 min) than by working manually (4.55e4.78 min). The mean difference was 4.36 min. Discussion Subjective methods of IQ evaluation are often used in QC for Xray fluoroscopy systems. However, the results obtained do provide
11 11 11
11 11 11
5.17 5.17 5.17 5.17
3.06 3.06 3.06 3.06
0.011 0.22
0.46
an estimate of observer detection efficiency. It can be difficult when comparing results obtained by different observers, particularly for a contrast-detail test.10 According to Vano et al.,8 the selection of ROI size and position is critical for evaluation of the HCSR parameter. Within this study the Fluoro-QC software has the advantage that subjective variations (ROI size and location) do not occur on the basis of the particular observer. In terms of optimization of signal detectability in digital imaging, FOM has been presented as a valuable parameter to relate IQ and dose per frame. As expected, Vano et al.11 showed the tendency for FOM to decrease in value as PMMA thickness increases. In this study, there were slight differences in FOM between manual and automatic modes. Overall, performing QC tests have become more timeconsuming.12 This study presented new software to accelerate this procedure in fluoroscopy. Three tests were performed using FluoroQC and results were obtained more quickly (0.26e0.43 min) than by working manually (4.55e4.78 min). The mean difference was 4.36 min. This difference may be explained in part because, in manual mode more steps are necessary to perform the test. In this study, ROI size (manually drawn) was user-dependent (Table 1). Likewise, an inexperienced user may apply ROIs extremely large or too small due to the subjectivity of manual mode. This may be directly reflected in the QC results and reproducibility of the test. A further advantage of Fluoro-QC relates to ROI, which always positioned in the same location and with the same size. However, this procedure has limitations relating to test tool placement, which must be imaged in the same place, without rotation, and at the same source-detector distance. Study limitations The limitations affecting this study related to our evaluation of only one C-arm angulation (a single radiographic projection) according to the protocol used. However, the impact of using other angulations should be taken into account in future research. In addition, this software was evaluated only on a single fluoroscopy systems and wider studies with other systems are warranted. Conclusion The Fluoro-QC software is a feasible, rapid and free method to evaluate image quality parameters for fluoroscopy. This software will help radiographers to perform regular IQ tests and establish baselines for monitoring. It can also support the auditing of radiological practices. Furthermore, the Fluoro-QC approach can assist health authorities with investigations into the IQ of fluoroscopy services according to guidelines for QC. We suggest that further research is needed to determine whether the Fluoro-QC is able to investigate image degradation over time and across a range of imaging systems.
Please cite this article in press as: Oliveira M, et al., Application of newly developed Fluoro-QC software for image quality evaluation in cardiac Xray systems, Radiography (2017), https://doi.org/10.1016/j.radi.2017.12.006
4
M. Oliveira et al. / Radiography xxx (2017) 1e4
Conflict of interest statement None.
4. 5.
Acknowledgements 6.
The authors thank the Federal Institute of Bahia, Brazil and Taparaca University, Chile for support in conducting of this study. The readers can obtain Fluoro-QC contacting the first author by email: [email protected]. References 1. Dimov A, Vassileva J. Quality control of the X-ray fluoroscopy equipment in the period of their clinical use e methods and first results. Natl Conf Biomed Phys Eng; Sofia (Bulgaria) 2004;36:9. Bulgaria: INIS-BGe0887. 2. Vano E, Ubeda C, Fernandez JM, Sanchez RM, Prieto C. Dose assessment during the commissioning of flat detector imaging systems for cardiology. Radiat Protect Dosim 2009;136:30e7. 3. Gislason-Lee AJ, Keeble C, Malkin CJ, et al. Impact of latest generation cardiac interventional X-ray equipment on patient image quality and radiation
7. 8.
9. 10.
11. 12.
dose for trans-catheter aortic valve implantations. Br J Radiol 2016;89, 20160269. Miracle AC, Mukherji SK. Conebeam CT of the head and neck, part 1: physical principles. AJNR Am J Neuroradiol 2009;30(6):1088e95. Duan X, Wang J, Leng S, Schmidt B, Allmendinger T, Grant K, et al. Electronic noise in CT detectors: impact on image noise and artifacts. AJR Am J Roentgenol 2013;201(4):W626e32. Inaba Y, Chida K, Kobayashi R, Zuguchi M. A cross-sectional study of the radiation dose and image quality of X-ray equipment used in IVR. J Appl Clin Med Phys 2016;17:391e401. Rasband W. ImageJ. Bethesda, Maryland, USA: U.S. National Institutes of Health; 1997e2012. Available at: http://imagej.nih.gov/ij/. [Accessed 5 May 2017]. Vano E, Ubeda C, Geiger B, Martinez LC, Balter S. Influence of image metrics when assessing image quality from a test object in cardiac X-ray systems. J Digit Imaging 2011;24:331e8. Available on: http://www.leedstestobjects.com/index.php/phantom/tor-18fg/. [last accessed November 2017]. Marshall NW. The practical application of signal detection theory to image quality assessment in x-ray image intensifier-TV fluoroscopy. Phys Med Biol 2001;46(6):1631. Vano E, Ubeda C, Leyton F, Miranda P. Radiation dose and image quality for paediatric interventional cardiology. Phys Med Biol 2008;53:4049e62. Thomadsen B. Critique of traditional quality assurance paradigm. Int J Radiat Oncol Biol Phys 2008;71:S166e9.
Please cite this article in press as: Oliveira M, et al., Application of newly developed Fluoro-QC software for image quality evaluation in cardiac Xray systems, Radiography (2017), https://doi.org/10.1016/j.radi.2017.12.006
Contents lists available at ScienceDirect
Radiography journal homepage: www.elsevier.com/locate/radi
Application of newly developed Fluoro-QC software for image quality evaluation in cardiac X-ray systems M. Oliveira a, *, G. Lopez a, P. Geambastiani b, C. Ubeda c a
Department of Heath Technology and Biology, Federal Institute of Bahia, Emídio dos Santos e s/n, Salvador, Bahia, Brazil Cardio Pulmonar Hospital, Av. Anita Garibaldi, 2199 - Garibaldi, Salvador, Bahia, Brazil c Medical Technology Department, Health Sciences Faculty, Tarapaca University, Av. 18 de septiembre, n 2222, Arica, Chile b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 3 August 2017 Received in revised form 9 December 2017 Accepted 13 December 2017 Available online xxx
Introduction: A quality assurance (QA) program is a valuable tool for the continuous production of optimal quality images. The aim of this paper is to assess a newly developed automatic computer software for image quality (IR) evaluation in fluoroscopy X-ray systems. Methods: Test object images were acquired using one fluoroscopy system, Siemens Axiom Artis model (Siemens AG, Medical Solutions Erlangen, Germany). The software was developed as an ImageJ plugin. Two image quality parameters were assessed: high-contrast spatial resolution (HCSR) and signal-tonoise ratio (SNR). The time between manual and automatic image quality assessment procedures were compared. The paired t-test was used to assess the data. p Values of less than 0.05 were considered significant. Results: The Fluoro-QC software generated faster IQ evaluation results (mean ¼ 0.31 ± 0.08 min) than manual procedure (mean ¼ 4.68 ± 0.09 min). The mean difference between techniques was 4.36 min. Discrepancies were identified in the region of interest (ROI) areas drawn manually with evidence of user dependence. The new software presented the results of two tests (HCSR ¼ 3.06, SNR ¼ 5.17) and also collected information from the DICOM header. Significant differences were not identified between manual and automatic measures of SNR (p value ¼ 0.22) and HCRS (p value ¼ 0.46). Conclusion: The Fluoro-QC software is a feasible, fast and free to use method for evaluating imaging quality parameters on fluoroscopy systems. © 2017 The College of Radiographers. Published by Elsevier Ltd. All rights reserved.
Keywords: Image quality Fluoroscopy Image metrics Quality control
Introduction A quality assurance (QA) program is a valuable tool for the continuous production of optimal quality images. Uncalibrated equipment could produce insufficient image quality (IQ) and lead to an increase in radiation dose to patients and staff.1 In some cases, maintaining an acceptable level of IQ in interventional fluoroscopy procedures requires a substantial increase in radiation dose per frame.2 IQ is currently assessed for contrast, resolution and artefacts. For an X-ray system, the signal-to-noise ratio (SNR) is directly proportional to the square root of the X-ray dose used to create the image.3 SNR is the ratio of useful information by the standard deviation (SD) of the grey level (noise) value.4 Noise is derived from * Corresponding author. E-mail addresses: [email protected] (M. Oliveira), [email protected] (G. Lopez), [email protected] (P. Geambastiani), [email protected] (C. Ubeda).
the quantum noise properties of X-ray photons and the electronic noise (unrelated to the number of photons detected) of the detection system.5 Both noise sources affect quality control test results. There are no detailed studies investigating the IQ optimization of the X-ray equipment used in interventional radiology.6 The aim of this paper is to assess a newly developed automatic computer software based method for image quality evaluation in fluoroscopy systems. Materials and methods Description of Fluoro-QC macro The Fluoro-QC macro code was developed for ImageJ to analyse IQ for the fluoroscopy technique. ImageJ7 is open-source software and thus it does not require a user licence. The program works independently of the operating system. Fluoro-QC works as a postprocessing tool, although the original image is not manipulated (Fig. 1).
https://doi.org/10.1016/j.radi.2017.12.006 1078-8174/© 2017 The College of Radiographers. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Oliveira M, et al., Application of newly developed Fluoro-QC software for image quality evaluation in cardiac Xray systems, Radiography (2017), https://doi.org/10.1016/j.radi.2017.12.006
2
M. Oliveira et al. / Radiography xxx (2017) 1e4
Figure 3. Experimental setup of the PMMA phantom and test object.
The macro creates four rectangular regions of interest (ROI) on the DICOM image acquired in cine mode. The numerical IQ evaluation is always performed on three consecutive images for each series located at sequential positions 10, 12 and 15, so as to avoid the stability problems likely to occur in the first images of the series. The ROIs are positioned in the same place as in the study performed by Vano et al.8 IQ is evaluated by analysing the lowcontrast circles and the high-contrast spatial resolution (HCSR) groups. The parameters evaluated are: SNR, HCSR and one figure of merit (FOM) (Fig. 2). To evaluate SNR, the following calculation is used:
Figure 1. Software workflow.
½BG ROI1 ffi SNR ¼ rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðSD2ROI þSD2BG Þ
(1)
2
where ROI1 is the mean value of the pixel inside circle number 3 and BG is the background mean value of the pixel placed near the low-contrast circles. The standard deviation is SDROI for ROI1, while SDBG is the standard deviation for BG. The following equation is used to numerically evaluate HCSR:
HCSR ¼ SD3 SD4
(2)
where SD3 is the standard deviation for the pixel content in the ROI3, inside the eighth group in the central grid of the pattern bar, and SD4 is the standard deviation for the pixel content in the ROI4
Figure 2. An illustration of the four ROI's automatically applied to specific places on the fluoroscopy image from the TOR 18 FG test object.
Figure 4. Comparison of timing requirements for the manual and automatic (FluoroQC) QC process.
Please cite this article in press as: Oliveira M, et al., Application of newly developed Fluoro-QC software for image quality evaluation in cardiac Xray systems, Radiography (2017), https://doi.org/10.1016/j.radi.2017.12.006
M. Oliveira et al. / Radiography xxx (2017) 1e4
3
Table 1 HCRS, SNR and FOM results according to manual and automatic ROIs. Manual
Automatic
Area (pixels2)
Area (pixels2)
ROI 1 ROI 2 ROI 3 ROI 4 SNR User “A” 18.9 User “B” 21.4 User “C” 16 Mean and standard deviation
21.8 26.8 10
18.4 22.8 8.9
7.1 23.4 13.4
5.27 5.62 5.25 5.38 ± 0.17
HCSR FOM 3 3.5 3.1 3.2 ± 0.22
ROI 1 ROI 2 ROI 3 ROI 4 SNR HCSR FOM
0.012 15 0.013 15 0.011 15
15 15 15
placed in the periphery of the high-contrast groups and representative of the noise in this area. FOM was evaluated and defined as:
FOM ¼
SNR2 DAP
p-Value manual p-Value manual and automatic and automatic mode for HCRS mode for SNR
(3)
where DAP is the dose area product obtained by the DICOM header. The user did not need to manually choose the frame number because it was always selected (automatically) by the frame representing the middle of the acquisition series. Evaluation of Fluoro-QC Three users with considerable experience in IQ evaluation calculated SNR and HCSR on the same image, both manually and using Fluoro-QC. A timer was used in order to compare the time consumed between the two procedures. Time consumed was counted from image selection until the presentation of results. Images were obtained for one X-ray system, the Siemens Axiom Artis model (Siemens AG, Medical Solutions Erlangen, Germany). Analysis was performed on images in the standard reduced format of 512 512 and 8 bits. Fig. 3 shows the experimental arrangement with the polymethyl methacrylate (PMMA) phantom and physical test object (Leeds TOR 18-FG).9 Statistical analysis Microsoft Office Excel software was used for statistical analysis and the paired t-test was used to assess the data. p values of less than 0.05 were considered significant. Results The results of the IQ assessment and the information from the DICOM header (manufacturer, the number of frames, tube voltage (kV), tube current (mA), exposure time(s), dose area product, intensifier size, matrix size, radiation mode and source-to-patient distance) were presented in a final report. Fig. 4 shows the time comparison between the automatic and manual tests. Significant differences were not demonstrated between the manual and automatic QC methods for SNR (p ¼ 0.22) and HCRS (p ¼ 0.46) values. The results of HCRS and SNR in Manual mode were influenced by discrepant values of ROIs size (Table 1). Three tests were performed using Fluoro-QC and results were obtained more quickly (0.26e0.43 min) than by working manually (4.55e4.78 min). The mean difference was 4.36 min. Discussion Subjective methods of IQ evaluation are often used in QC for Xray fluoroscopy systems. However, the results obtained do provide
11 11 11
11 11 11
5.17 5.17 5.17 5.17
3.06 3.06 3.06 3.06
0.011 0.22
0.46
an estimate of observer detection efficiency. It can be difficult when comparing results obtained by different observers, particularly for a contrast-detail test.10 According to Vano et al.,8 the selection of ROI size and position is critical for evaluation of the HCSR parameter. Within this study the Fluoro-QC software has the advantage that subjective variations (ROI size and location) do not occur on the basis of the particular observer. In terms of optimization of signal detectability in digital imaging, FOM has been presented as a valuable parameter to relate IQ and dose per frame. As expected, Vano et al.11 showed the tendency for FOM to decrease in value as PMMA thickness increases. In this study, there were slight differences in FOM between manual and automatic modes. Overall, performing QC tests have become more timeconsuming.12 This study presented new software to accelerate this procedure in fluoroscopy. Three tests were performed using FluoroQC and results were obtained more quickly (0.26e0.43 min) than by working manually (4.55e4.78 min). The mean difference was 4.36 min. This difference may be explained in part because, in manual mode more steps are necessary to perform the test. In this study, ROI size (manually drawn) was user-dependent (Table 1). Likewise, an inexperienced user may apply ROIs extremely large or too small due to the subjectivity of manual mode. This may be directly reflected in the QC results and reproducibility of the test. A further advantage of Fluoro-QC relates to ROI, which always positioned in the same location and with the same size. However, this procedure has limitations relating to test tool placement, which must be imaged in the same place, without rotation, and at the same source-detector distance. Study limitations The limitations affecting this study related to our evaluation of only one C-arm angulation (a single radiographic projection) according to the protocol used. However, the impact of using other angulations should be taken into account in future research. In addition, this software was evaluated only on a single fluoroscopy systems and wider studies with other systems are warranted. Conclusion The Fluoro-QC software is a feasible, rapid and free method to evaluate image quality parameters for fluoroscopy. This software will help radiographers to perform regular IQ tests and establish baselines for monitoring. It can also support the auditing of radiological practices. Furthermore, the Fluoro-QC approach can assist health authorities with investigations into the IQ of fluoroscopy services according to guidelines for QC. We suggest that further research is needed to determine whether the Fluoro-QC is able to investigate image degradation over time and across a range of imaging systems.
Please cite this article in press as: Oliveira M, et al., Application of newly developed Fluoro-QC software for image quality evaluation in cardiac Xray systems, Radiography (2017), https://doi.org/10.1016/j.radi.2017.12.006
4
M. Oliveira et al. / Radiography xxx (2017) 1e4
Conflict of interest statement None.
4. 5.
Acknowledgements 6.
The authors thank the Federal Institute of Bahia, Brazil and Taparaca University, Chile for support in conducting of this study. The readers can obtain Fluoro-QC contacting the first author by email: [email protected]. References 1. Dimov A, Vassileva J. Quality control of the X-ray fluoroscopy equipment in the period of their clinical use e methods and first results. Natl Conf Biomed Phys Eng; Sofia (Bulgaria) 2004;36:9. Bulgaria: INIS-BGe0887. 2. Vano E, Ubeda C, Fernandez JM, Sanchez RM, Prieto C. Dose assessment during the commissioning of flat detector imaging systems for cardiology. Radiat Protect Dosim 2009;136:30e7. 3. Gislason-Lee AJ, Keeble C, Malkin CJ, et al. Impact of latest generation cardiac interventional X-ray equipment on patient image quality and radiation
7. 8.
9. 10.
11. 12.
dose for trans-catheter aortic valve implantations. Br J Radiol 2016;89, 20160269. Miracle AC, Mukherji SK. Conebeam CT of the head and neck, part 1: physical principles. AJNR Am J Neuroradiol 2009;30(6):1088e95. Duan X, Wang J, Leng S, Schmidt B, Allmendinger T, Grant K, et al. Electronic noise in CT detectors: impact on image noise and artifacts. AJR Am J Roentgenol 2013;201(4):W626e32. Inaba Y, Chida K, Kobayashi R, Zuguchi M. A cross-sectional study of the radiation dose and image quality of X-ray equipment used in IVR. J Appl Clin Med Phys 2016;17:391e401. Rasband W. ImageJ. Bethesda, Maryland, USA: U.S. National Institutes of Health; 1997e2012. Available at: http://imagej.nih.gov/ij/. [Accessed 5 May 2017]. Vano E, Ubeda C, Geiger B, Martinez LC, Balter S. Influence of image metrics when assessing image quality from a test object in cardiac X-ray systems. J Digit Imaging 2011;24:331e8. Available on: http://www.leedstestobjects.com/index.php/phantom/tor-18fg/. [last accessed November 2017]. Marshall NW. The practical application of signal detection theory to image quality assessment in x-ray image intensifier-TV fluoroscopy. Phys Med Biol 2001;46(6):1631. Vano E, Ubeda C, Leyton F, Miranda P. Radiation dose and image quality for paediatric interventional cardiology. Phys Med Biol 2008;53:4049e62. Thomadsen B. Critique of traditional quality assurance paradigm. Int J Radiat Oncol Biol Phys 2008;71:S166e9.
Please cite this article in press as: Oliveira M, et al., Application of newly developed Fluoro-QC software for image quality evaluation in cardiac Xray systems, Radiography (2017), https://doi.org/10.1016/j.radi.2017.12.006