thesis research
Latif
NewVersionofthethesisreserch.docxEstimation of Peak Skin Dose and Its Relation to the Size Specific Dose Estimate
Abdullatif Abdullah1, *, Matthew Williams 2, *
13700 O St NW, Washington, DC, 20057
23800 Reservoir Rd NW, Washington, DC, 20007
Received December 10, 2021, amended month date year, accepted month date year
Abstract:
The research aims to test the concept that the size specific dose estimete (SSDE) has a significantly higher linear relationship with the peak skin dose (PSD). To determine the connection between the two measurements, a NEMA phantom and ACR phantom were provided with a peak skin dose measured using the external dosimeters (Nanodots) to measure the level of delivered radiation. For the experiment's success, different measuring techniques and methods were utilized in the research. Nanodots dosimeters which use optically stimulated luminescence (OSL) technology, were used to determine peak skin dose. The findings revealed a somewhat favorable connection in both PA and lateral regions, suggesting that the PSD and SSDE may be related. Posterior and Lateral angles have some potential since, in most projections, the greater the PSD, the higher the SSDE. The measured PSD and SSDE revealed that a physicist can predict the PSD within 80% of the actual dose estimates with considerable some uncertainties.
1
2
3
Introduction
The Federal Drug Administration (FDA) states that the Peak Skin Dose is the highest radiation dose accruing actually at a single site on a patient’s skin (1). Knowing the appropriate highest dosage is vital so that no harm is caused to the patient. The United States has regulated that the fluoroscopic system provides a display of the irradiation time, dose rate at the interventional reference point during irradiation, and the cumulative dose for the procedure upon completion of irradiation (2). In preparation for actual patients, technologists and physicists would revert to the manufactured dose estimation which is called the Computed Tomography Dose Index (CTDI). The CTDI is generally utilized for quality control including the radiation output of CT machines (2). Specifically, the volume CTDI is shown on the control center of all CT machines and is promptly accessible to the administrator. In any case, the CT Dose Index (CTDIvol) was originally designed as an index of dose associated with various CT diagnostic procedures, not as a direct dosimetry method for individual patient dose assessments (3).
Moreover, CTDIvol is reported in two units: a 16-cm phantom for head exams or 32-cm phantom or body exams. The relationship between the CTDIvol and skin dose entrance depends on various factors, one of which is the patient size (5). CTDIvol is displayed on the console of CT scanners, and it gives genuine estimates of the dose being delivered to a phantom,
which can over or under estimate dose to an actual patient. For instance, report 204 was generated to help better estimate the dose to a patient based on the patient size and shape that differs from size and shape of phantoms used to generate CTDIvol (4).
CTDIvol is a measurement taking with an ion chamber inserted into a phantom but not laying on outside of a phantom at the point of skin entrance. Specifically, estimating Peak Skin Dose is ideal since it is a surely known dosimetric amount that directly identifies with radiation-incited skin injures. For example, radiation directly damages the skin as well as its deep tissue cells, causing dryness, loss of elasticity, pigmentation, soft tissue fibrosis, capillary dilatation, and radiation dermatitis in irradiated areas. Therefore, there is a threshold range for transient erythema and temporary epilation is 2-5 Gray (Gy) peak skin dose, prolonged erythema and permanent partial epilation have threshold ranges of 5-10 Gy peak skin dose and severe skin injury is associated with larger values of peak skin dose (5).
Besides, estimates of PSD values, utilizing appropriate phantoms can be made across all types of CT units and scan protocols accessible in clinics (6). This is significant for comparing doses for a similar CT examination in different facilities, which can change fundamentally. More recently, modifications to the original CTDI concept have attempted to convert it into to patient dosimetry method, but have mixed results in terms of accuracy (7).
Nonetheless, CTDI-based dosimetry is the current worldwide standard for estimation of patient dose in CT (8). Therefore, CTDIvol is often used to enable medical physicists to compare the dose output between different CT scanners (8). Also, since CTDIvol estimates the patient's radiation exposure from the CT procedure, the exposures are the same regardless of
patient size, but the size of the patients is a factor in the overall patient's absorbed dose (SSDE) (4). The size-specific dose estimate (SSDE) is measured in mGy, and it is a method of estimating CT radiation dose that takes a patient's size into account (4).
From a radiation protection point of view, determining the maximum dose delivered to the skin would allow deriving quantities that can be compared with dose reference levels set by national and international standards. The most important outcome from a radiation safety perspective is evaluating if a radiation injury had occurred quickly (10). In this research, the peak skin dose delivered to a patient was estimated experimentally by measuring the dose delivered to the surface of a NEMA phantom and an ACR phantom using external dosimeters (11). These dosimeters provided PSD values for a given protocol and its related CTDIvol. From this, a relationship was evaluated between both quantities. The aim of this project was to test the hypothesis that the size-specific dose estimate (SSDE) has a sufficiently strong linear relationship with PSD to allow direct calculation of the PSD directly from the SSDE. Comment by ,Abdullatif abdullah: Prof. Jorgensen note “What are you saying here? Do you mean that dose rate is the most important parameter determining skin injury? Or do you mean that shorter latency means higher doses? I don’t understand why this sentence is stuck in here.”
MATERIALS AND METHODS
The measurements were performed with a Siemens 64 slices, Biograph mCT. A comparison was made between the CTDIvol value displayed on the CT console and the measured CTDIvol value using the AAPM protocol. For every examined scanner, the CTDIvol was obtained from scans in an axial mode for head scans and helical mode of the routine pelvis, cervical spine, abdomen, and thoracic scans using the scan parameters. The corresponding CTDIvol displayed on the console was recorded as shown in Table 1.
Peak Skin Dose was estimated by using Nanodots dosimeters (12) (International Specialty Products, Inc., Wayne, NJ, USA) which have optically stimulated luminescence (OSL) technology which is a single point radiation monitoring dosimeter. It is a useful tool in measuring the patient dose, and it is an ideal solution in multiple settings, including diagnostic radiology, nuclear medicine, interventional procedures and radiation oncology (12).
Nanodots dosimeters also have minimal angular or energy dependencies with appropriate calibration which can be used to measure skin dose at a point of interest. Moreover, LANDAUER provides a set of calibration dosimeters exposed at a beam quality of 80 kVp on a PMMA phantom at normal incidence for conventional (non-mammography) diagnostic radiology applications (12). For radiation oncology applications, LANDAUER provides a set of screened, unexposed calibration dosimeters that can be irradiated using a radiation therapy beam quality. Another way for calibration is to request a dosimeter set exposed to a 662 keV beam quality (Cs-137) (12).
The Nanodot dosimeters were placed on three different locations (Anterior-Posterior, Lateral (LAT) and Posterior-Anterior) as shown in figure 1, and the dose to the skin was measured at these locations.
1
2
3
CT TABLE
Figure1. (a) a set-up used for the Peak Skin Dose measurements with placing the Nanodots on the three locations (Anterior, Posterior and lateral) with using the NEMA phantom. The cylindrical phantom was placed in the center of the CT scan. (b) The Peak Skin Dose measurements with using the ACR phantom which was placed in the center of the CT scan. Comment by ,Abdullatif abdullah: I will change this figure once I am back on campus. I will place both phantoms we used in the center of the scanner and upload the pictures here instead.
Experimental set-up and procedure:
The CTDIvol displayed by the scanner was validated to the true CTDIvol following the ACR testing guidelines (13). A correction factor was used to correct the inaccuracies in the displayed value. This correction was applied to the DLP displayed by the scanner.
Peak skin dose and its relation were measured by the 2 phantoms, and the phantoms were aligned at the isocenter of the scanner and a single axial CT scan was made. After placing the Nanodot dosimeters on the AP, LAT and PA locations, the phantoms were scanned over the scan length for a fixed value of the tube current. The measurement was repeated several times using various scanning techniques (with varying energy, current) as shown in table 1. Size conversion factors used were based on the dimension of the phantom being scanned. These K-factors with the CTDIvol produced the size-specific dose estimates (SSDEs), and since the CT dose index was provided at the CT scanner too, the size-specific dose estimate for the phantoms was calculated. Also testing if the correlation between the size-specific dose estimate and the measurement of the peak skin dose match was done, and since such a relationship exists, finding that factor was achieved.
RESULTS
To determine the connection between the Peak Skin Dose and Size Specific Dose Estimates, a comparison was done using the NEMA phantom and ACR phantom with a peak skin dose measured using the Nanodots. After testing if the correlation between the size-specific dose estimate and the measurement of the peak skin dose, the SSDE was calculated using the corresponding k-factor based on the AP and lateral dimension from AAPM Report 204. The Size Specific Dose estimate values as shown in table 1, are based on the CTDIvol value which was displayed on the console. (effective diameter = ), the effective diameter represented the diameter of the phantom at a given location along the z-axis of the phantom (4). The anterior (AP) in the NEMA phantom was 22.37 cm whereas the lateral (LAT) in the same phantom was 29.27 cm. The effective diameter for the NAMA phantom was 26 cm.
(SSDE = CTDIvol x K factor), this formula was used to solve for the Size Specific Dose Estimates. Report 204, has tables with different conversion factors based on the use of the NEMA phantom. Table 1D in the report, provided conversion factors as a function of the
Tables 1 illustrates the difference kilovoltage peaks, milliampere-seconds, CTDIvol in mGy, the Peak Skin Dose of all the Nanodot dosimeters that were used in the specific locations and the measured size specific dose estimate.
effective diameter. The data on the report are used when the CTDIvol is known. In this study, the conversion factor for 26 cm effective diameter NEMA phantom in all the projections besides head was 1.43. For head scans, different protocol was used which was similar to the protocol used with the ACR phantom. The effective diameter for the ACR phantom was 19 cm. The AAPM Report 204 stated that the conversion
factor based on the use of the ACR phantom with 19 cm effective diameter was 0.90. The SSDE was calculated and measured in mGy as shown in table 1.
Table 1 includes a comparison between the peak skin dose that was measured by the Nanodots in the three locations. The measured PSD values varied between 0.13 and 70.77 mGy. The highest PSD was measured using the ACR phantom with head projection at the anterior location and the lowest was Peak Skin Dose was measured at the anterior location of the NEMA phantom with abdomen projection.
Figure 2: The graph illustrates the relationship between Peak Skin Dose in AP location and the Size Specific Dose Estimates in AP location in the NEMA phantom and the ACR phantom.
Figure 2, illustrates the measured PSD in AP location against the SSDE in AP location with using 2 different phantoms (NEMA phantom and ACR phantom). For both phantoms, there was a linear relationship between the size specific dose estimates and the peak skin dose. In this study, the R-squared value was 0.75 which indicate that 75% of the variance of the dependent variable being studied is explained by the variance of the independent variable (14). Therefore, the relationship between the PSD in AP location and the SSDE in AP location has a positive correlation.
Figure 3: The graph demonstrates the relationship between Peak Skin Dose in PA location and the Size Specific Dose Estimates in PA location in the NEMA phantom and the ACR phantom.
The third figure demonstrates the measured PSD in PA location against the SSDE in PA location. For both phantoms, there was linear relationship between the size specific dose estimates and the peak skin dose. In this graph, the R-squared value was 0.85. Therefore, the relationship between the PSD in PA location and the SSDE has a positive relationship and high correlation.
Figure 4: The graph illustrates the relationship between Peak Skin Dose in Lateral location and the Size Specific Dose Estimates in Lateral location in the NEMA phantom and the ACR phantom.
Figure 4, illustrates the measured PSD in the lateral location against the SSDE in Lateral location. For both phantoms, there was linear relationship between the
size specific dose estimates and the peak skin dose. In this graph the R-squared value was 0.78 which indicated that there was a positive linear relationship between the PSD in lateral location and the SSDE in lateral location.
In all the plots, linear relationship between the PSD and SSDE was found, and the linear fitting equation was calculated. Peak Skin Dose can be predicted in the anterior location with knowing the CTDIvol which is shown on the console. (SSDE = 0.8373 x (PSD) - 2.3412), this was the fitting equation for the AP location graph (1st graph).
(SSDE = 0.6833 x (PSD) + 4.9856), this was the linear fitting equation for the posterior location graph (2nd graph) which indicates that PSD can be measured with knowing the SSDE value.
(SSDE = 0.9384 x (PSD) - 3.084), this was the linear fitting equation for the lateral location graph (3rd graph).
The three equations have a high positive relationship, so predicting the value of SSDE or PSD will be possible but not 100% accurate. With using these data and fitting equations, a physicist can estimate the PSD, but with some limitations.
The physicist would be within 80% the true dose estimates and a large error would be there as well. The regression was almost 80% in the three locations, so roughly 80% of the data points will fall close to the linear line.
DISCUSSION
The regression of the Peak Skin Dose was different in the AP and LAT locations comparison with the lateral location which is because the thickness of the phantom. Considering that examination is performed in the lateral location of the body which has the highest x-ray attenuation, thus requiring higher beam energy to penetrate. With increasing the patient average diameter, the peak skin dose was higher. According to the data that was measured, the measured PSD was higher in all the lateral location than the AP and PA locations. The bigger the phantom (more tissue to penetrate), the more dose was required to attenuate and reached the dosimeter.
In the is study the AP, PA and lateral dimensions of the phantom were used to measure the SSDE which is a factor that is used to estimate the absorbed dose. This could’ve been an error in measuring the peak skin dose since the SSDE was not measured at that time. Also, there was a linear relationship between the PSD and the SSDE because the Size Specific Dose Estimates dictate the patient’s dose and this could be one of the reasons that the linear relationship occurred. Also, there could be better modifications to the K-factors in order to dictate the patient’s more accurately.
When calculating how much radiation dose a patient is actually receiving, it’s best to consider their actual size. CTDIvol and DLP are common methods to estimate a patient's radiation dose from a CT procedure. The dose is the same regardless of patient size, but the size of the patients is a factor in the overall patient's absorbed dose. Therefore, SSDE measured in mGy, would allow the physicists to use the patient’s size as a factor in order to estimate the radiation dose. In the other hand the PSD is the maximum absorbed dose in mGy to the most heavily exposed region of the skin in specific location. In this study, the measured values of the PSD and SSDE had a linear relationship in most projections (C-spine, thoracic and pelvis). The higher the PSD was, the higher the SSDE which was due to the measured CTDIvol which displayed in the console (the higher the CTDIvol was, the higher SSDE was calculated).
There is a difference between the CTDIvol that was shown on the console and the actual CTDIvol. The CTDIvol or its derivative the DLP, as seen on consoles and outputted, do not represent the actual absorbed or effective dose for the patient. They should be taken as an index of radiation output by the system for comparison purposes. In this study, it is not possible to compare the true CTDIvol to the displayed because the phantoms that were used were not CTDI phantoms, so it is not possible to place a CTDI probe.
However, nowadays many modifications to original CTDI concept have attempted to make it more accurate patient dosimetry method, with mixed results. Body CTDIvol reported by the CT scanner, or measured on a CT scanner, is a dose index that results from air kerma measurements at two locations, to a very cylinder of plastic phantom with a density of 1.19 g/cm3 (15).
CTDIvol is primary affected by kVp and mAs, so parameters which were used in most of the previous researches such as (Auto-kVp and dose reduction initiative parameters) were not used in this research because in a normal system as the tube rotate around the phantom, the thickness of the phantom varies, LAT location usually requires more dose so more kVp and mAs than AP location, so the kVp and mAs values will fluctuate which will lead to change the CTDIvol as well. Therefore, fixed kVp and mAs values were used in order to have a continuous CTDIvol for the entire process, and for a different phantom a different parameter was used in different exam so an abdomen phantom had different parameter than the one for the cervical spine.
According to the measured data, some scan projections such as abdomen had high PSD and high SSDE due to the high measured CTDIvol and DLP caused out wire and low regression. Taking out the abdomen PSD and SSDE from the graphs make the regression higher (more positive) which means correlation could exist. Therefore, some projections such as an abdomen and head might make the data points and graphs not clear and hard to be read.
When graphing the measured PSD and SSDE in each phantom separately, a higher regression (more positive correlation) was found (close to 90%) in all the three locations. This means that the closer the patient to become cylindrical, the better relationship between PSD and SSDE will be and more accurate doses will be measured. It fails at very large effective circumferences with perfectly cylindrical patients.
In this study, only two phantoms were used (NEMA and ACR phantoms) with specific thicknesses, so other phantoms such as anthropomorphic phantoms and fake human phantoms with different thickness styles could be used to get better data and correlation.
In this study, only 9 measurements were taken in the three different location due to the limitation of the Nanodots. More measurements could have been taken and a better data points would have been measured. With more data testing that the SSDE has a sufficiently strong linear relationship with PSD could be proven.
CONCLUSION
REFERENCES
Center for Devices and Radiological Health. (n.d.). Radiation Dose Quality Assurance: Questions and Answers. U.S. Food and Drug Administration. Retrieved November 19, 2021, from https://www.fda.gov/radiation-emitting-products/initiative-reduce-unnecessary-radiation-exposure-medical-imaging/radiation-dose-quality- assurance-questions-and-answers.
De las Heras, H., Minniti, R., Wilson, S., Mitchell, C., Skopec, M., Brunner, C. C., & Chakrabarti, K. (2013). Experimental estimates of peak skin dose and its relationship to the CT dose index using the CTDI head phantom. Radiation protection dosimetry, 157(4), 536-542.
Jones, A. K., Kisiel, M. E., Rong, X. J., & Tam, A. L. (2021). Validation of a method for estimating peak skin dose from CT‐guided procedures. Journal of applied clinical medical physics.
Publications. AAPM Publications - AAPM Reports. (2021). Retrieved 17 December 2021, from https://www.aapm.org/pubs/reports/rpt_204.pdf
Stephen Balter et al., “Fluoroscopically Guided Interventional Procedures: A Review of Radiation Effects on Patients’ Skin and Hair,” Radiology Vol. 254, No. 2, pp. 326-341, February 2010.
Tack, D., & Gevenois, P. A. (2018). Radiation dose from adult and pediatric Multidetector computed tomography. Springer Science & Business Media.
McCollough, C. H., Leng, S., Yu, L., Cody, D. D., Boone, J. M. and McNitt-Gray, M. F. CT dose index and patient dose: they are not the same thing. Radiology 259(2), 311 – 316 (2011).
Bauhs, J. A., Vrieze, T. J., Primak, A. N., Bruesewitz, M. R. and McCollough, C. H. CT dosimetry: Comparison of measurement techniques and devices. Radiographics 28, 245 – 253 (2008).
Nationals Council on Radiation Protection and measurements. NCRP Report No. 160: Ionizing radiation exposure of the population of the United States (2009).
Beganovic, A., Sefic-Pasic, I., Skopljak-Beganovic, A., Kristic, S., Sunjic, S., Mekic, A., Gazdic-Santic, M., Drljevic, A. and Samek, D. Doses to skin during dynamic perfusion computed tomography of the liver. Radiat. Prot. Dosim. 153(1), 106–111 (2013).
NanoDot™. LANDAUER. (n.d.). Retrieved November 19, 2021, from https://www.landauer.com/product/nanodot.
ACR–sar–SPR practice parameter for the performance of ... (n.d.). Retrieved November 19, 2021, from https://www.acr.org/-/media/ACR/Files/Practice-Parameters/CT-Entero.pdf.
Frost, J. (2021). How To Interpret R-squared in Regression Analysis. Retrieved 1 December 2021, from https://statisticsbyjim.com/regression/interpret-r-squared-regression/
Morgan, M. (2021). Size specific dose estimate | Radiology Reference Article | Radiopaedia.org. Retrieved 1 December 2021, from https://radiopaedia.org/articles/size-specific-dose-estimate?lang=us
McCollough, C., Leng, S., Yu, L., Cody, D., Boone, J., & McNitt-Gray, M. (2011). CT Dose Index and Patient Dose: They AreNotthe Same Thing. Radiology, 259(2), 311-316. doi: 10.1148/radiol.11101800
NEMA phantom 12.53 7.95 12.55 34.119999999999997 1.85 3.43 3.39 16.41 78.92 0.39 11.69 12.55 32.270000000000003 4.88 7.12 5.94 30.54 56.43 ACR phantom 1.85 3.43 3.39 16.41 78.92 4.88 7.12 5.94 30.54 56.43
Peak Skin Dose (mGy)
Size Specific Dose Estimates (mGy)
NEMA phantom 12.53 7.95 12.55 34.119999999999997 1.85 3.43 3.39 16.41 78.92 0.13 9.52 9.8800000000000008 27.33 ACR phantom 12.53 7.95 12.55 34.119999999999997 1.85 3.43 3.39 16.41 78.92 6.21 7.5 27.83 36.86 70.77
Peak Skin Dose (mGy)
Size Specific Dose Estimates (mGy)
NEMA phantom 12.53 7.95 12.55 34.119999999999997 1.85 3.43 3.39 16.41 78.92 0.19 9.64 10.8 30.05 ACR phantpm 12.53 7.95 12.55 34.119999999999997 1.85 3.43 3.39 16.41 78.92 5.5 5.97 6.17 27.16 62.76
Peak Skin Dose (mGy)
Size Specific Dose Estimates (mGy)
