A machine learning approach for correcting glow curve anomalies in CaSO4:Dy-based TLD dosimeters used in personnel monitoring.

The study presents a novel approach to analysing the thermoluminescence (TL) glow curves (GCs) of CaSO4:Dy-based personnel monitoring dosimeters using machine learning (ML). This study demonstrates the qualitative and quantitative impact of different types of anomalies on the TL signal and trains ML algorithms to estimate correction factors (CFs) to account for these anomalies. The results show a good degree of agreement between the predicted and actual CFs, with a coefficient of determination greater than 0.95, a root mean square error less than 0.025, and a mean absolute error less than 0.015. The use of ML algorithms leads to a significant two-fold reduction in the coefficient of variation of TL counts from anomalous GCs. This study proposes a promising approach to address anomalies caused by dosimeter, reader, and handling-related factors. Furthermore, it accounts for non-radiation-induced TL at low dose levels towards improving the dosimetric accuracy in personnel monitoring.

Monte Carlo-based dosimetric studies of a locally developed170Tm LDR brachytherapy seed source.

170Tm is being explored as a source for applications in brachytherapy. Although it has adequate physical properties, such as a short half-life (128.6 d), high specific activity and a mean photon energy of about 66 keV, it has a drawback of low photon yield (only about six photon emissions/100 beta emissions). The objective of this work is to study the dosimetric characteristics of a locally developed170Tm brachytherapy seed source using the Monte Carlo-based EGSnrc code system. In this study, we calculate the dose rate constant, air-kerma strength, radial dose function, anisotropic function and 2D dose-rate distributions in water. Separate simulations are carried out by considering the photon (gamma and characteristic x-ray) and beta spectra of the source. For regions close to the source (surface of the source

Dosimetry of indigenously developed 177Lu patch source for surface brachytherapy-Experimental and Monte Carlo methods.

This paper describes the evaluation of dosimetry characteristics of an in-house developed 177Lu skin patch source for treatment of non-melanoma skin cancer. A 177Lu skin patch source based on Nafion-115 membrane backbone containing 3.46 ± 0.01 mCi of activity was used. Activity measurement of the patch source was based on gamma ray spectrometry using a HPGe detector. The efficiencies of the HPGe detector were fitted using an orthogonal polynomial function. The absorbed dose rate to water at 5 µm depth in water was determined using an extrapolation chamber, EBT3 Gafchromic film and compared with Monte Carlo methods. The correction factors such as Bragg-Gray stopping power ratio of water-to-air and chamber wall material being different from water, needed to be applied on measurements for establishing the dose rate at 5 µm depth, were calculated using the Monte Carlo method. Absorbed dose rate at 5 µm depth in water (surface dose rate) measured using an extrapolation chamber and EBT3 Gafchromic film were 9.9 ± 0.7 and 8.2 ± 0.1 Gy h-1 mCi-1 respectively for the source activity of 3.46 ± 0.01 mCi. The surface dose rate calculated using the Monte Carlo method was 8.7 ± 0.2 Gy h-1 mCi-1, which agrees reasonably well with measurement. The measured dose rate per mCi offers scope for ascertaining treatment time required to deliver the dose for propitious therapeutic outcome. Additionally, on-axis depth dose and lateral dose profiles at 5 µm and 1 mm depth in water phantom were also calculated using the Monte Carlo method.

Monte Carlo-based investigation of absorbed-dose energy dependence of radiochromic films in high energy brachytherapy dosimetry.

Relative absorbed dose energy response correction, R, for various radiochromic films in water phantom is calculated by the use of the Monte Carlo-based EGSnrc code system for high energy brachytherapy sources 60Co, 137Cs, 192Ir and 169Yb. The corrections are calculated along the transverse axis of the sources (1-15 cm). The radiochromic films investigated are EBT, EBT2 (lot 020609 and lot 031109), RTQA, XRT, XRQA, and HS. For the 60Co source, the value of R is about unity and is independent of distance in the water phantom for films other than XRT and XRQA. The XRT and XRQA films showed distance dependent R values for this source (the values of R at 15 cm from the source in water are 1.845 and 2.495 for the films XRT and XRQA, respectively). In the case of 137Cs and 192Ir sources, XRT, XRQA, EBT2 (lot 031109), and HS films showed distant-dependent R values. The rest of the films showed no energy dependence (HS film showed R values less than unity by about 5%, whereas the other films showed R values higher than unity). In the case of 169Yb source, the EBT film showed no energy dependence and EBT2 film (lot 031109) showed a distance-independent R value of 1.041. The rest of the films showed distance-dependent R values (increases with distance for the films other than HS). The solid phantoms PMMA and polystyrene enhance the R values for some films when compared the same in the water phantom.

Monte Carlo-based beam quality and phantom scatter corrections for solid-state detectors in 60Co and 192Ir brachytherapy dosimetry.

Beam quality correction, kQQ0(r), for solid-state detectors diamond, LiF, Li2B4O7, Al2O3, and plastic scintillator are calculated as a function of distance, r, along the transverse axis of the 60Co and 192Ir brachytherapy sources using the Monte Carlo- based EGSnrc code system. This study also includes calculation of detector-specific phantom scatter correction, kphan(r), for solid phantoms such as PMMA, polysty- rene, solid water, virtual water, plastic water, RW1, RW3, A150, and WE210. For 60Co source, kQQ0(r) is about unity and distance-independent for diamond, plastic scintillator, Li2B4O7 and LiF detectors. For this source, kQQ0(r) decreases gradually with r for Al2O3 detector (about 6% smaller than unity at 15 cm). For 192Ir source, kQQ0(r) is about unity and distance-independent for Li2B4O7 detector (overall variation is about 1% in the distance range of 1-15 cm). For this source, kQQ0(r) increases with r for diamond and plastic scintillator (about 6% and 8% larger than unity at 15 cm, respectively). Whereas kQQ0(r) decreases with r gradually for LiF (about 4% smaller than unity at 15 cm) and steeply for Al2O3 (about 25% smaller than unity at 15 cm). For 60Co source, solid water, virtual water, RW1, RW3, and WE210 phantoms are water-equivalent for all the investigated solid-state detectors. Whereas polystyrene and plastic water phantoms are water-equivalent for diamond, plastic scintillator, Li2B4O7 and LiF detectors, but show distance-dependent kphan(r) values for Al2O3 detector. PMMA phantom is water-equivalent at all distances for Al2O3 detector, but shows distance-dependent kphan(r) values for remaining detec- tors. A150 phantom shows distance-dependent kphan(r) values for all the investigated detector materials. For 192Ir source, solid water, virtual water, RW3, and WE210 phantoms are water-equivalent for diamond, plastic scintillator, Li2B4O7 and LiF detectors, but show distance-dependent kphan(r) values for Al2O3 detector. All other phantoms show distance-dependent kphan(r) values for all the detector materials. 

Monte Carlo calculation of beam quality correction for solid-state detectors and phantom scatter correction at 137Cs energy.

Beam quality correction kQQo (r), which reflects the absorbed energy dependence of the detector, is calculated for solid state detector materials diamond, LiF, Li2B4O7 and Al2O3 for the 137Cs RTR brachytherapy source using the Monte Carlo-based EGSnrc code system. The study also includes calculation of detector-specific phantom scatter corrections kphant (r) for solid phantoms such as PMMA, polystyrene, RW1, solid water, virtual water and plastic water. Above corrections are calculated as a function of distance r along the transverse axis of the source. kQQo (r) is about unity for the Li2B4O7 detector. LiF detector shows a gradual decrease in kQQo (r) with r (decrease is about 2% over the distance range of 1 - 15 cm). Diamond detector shows a gradual increase in kQQo (r)with r (about 3% larger than unity at 15 cm). In the case of Al2O3 detector, kQQo (r)decreases with r steeply (about 14% over the distance range of 1 - 15 cm). The study shows that some solid state detectors demonstrate distance-dependent kphant (r)values, but the degree deviation from unity depends on the type of solid phantom and the detector.

Calculation of biological effectiveness of SOBP proton beams: a TOPAS Monte Carlo study.

Objective.This study aims to investigate the biological effectiveness of Spread-Out Bragg-Peak (SOBP) proton beams with initial kinetic energies 50-250 MeV at different depths in water using TOPAS Monte Carlo code.Approach.The study modelled SOBP proton beams using TOPAS time feature. Various LET-based models and Repair-Misrepair-Fixation model were employed to calculate Relative Biological Effectiveness (RBE) for V79 cell lines at different on-axis depths based on TOPAS. Microdosimetric Kinetic Model and biological weighting function-based models, which utilize microdosimetric distributions, were also used to estimate the RBE. A phase-space-based method was adopted for calculating microdosimetric distributions.Main results.The trend of variation of RBE with depth is similar in all the RBE models, but the absolute RBE values vary based on the calculation models. RBE sharply increases at the distal edge of SOBP proton beams. In the entrance region of all the proton beams, RBE values at 4 Gy i.e. RBE(4 Gy) resulting from different models are in the range of 1.04-1.07, comparable to clinically used generic RBE of 1.1. Moving from the proximal to distal end of the SOBP, RBE(4 Gy) is in the range of 1.15-1.33, 1.13-1.21, 1.11-1.17, 1.13-1.18 and 1.17-1.21, respectively for 50, 100, 150, 200 and 250 MeV SOBP beams, whereas at the distal dose fall-off region, these values are 1.68, 1.53, 1.44, 1.42 and 1.40, respectively.Significance.The study emphasises application of depth-, dose- and energy- dependent RBE values in clinical application of proton beams.

Radiation-induced DNA damage by proton, helium and carbon ions in human fibroblast cell: Geant4-DNA and MCDS-based study.

Background. Radiation-induced DNA damages such as Single Strand Break (SSB), Double Strand Break (DSB) and Complex DSB (cDSB) are critical aspects of radiobiology with implications in radiotherapy and radiation protection applications.Materials and Methods. This study presents a thorough investigation into the effects of protons (0.1-100 MeV/u), helium ions (0.13-100 MeV/u) and carbon ions (0.5-480 MeV/u) on DNA of human fibroblast cells using Geant4-DNA track structure code coupled with DBSCAN algorithm and Monte Carlo Damage Simulations (MCDS) code. Geant4-DNA-based simulations consider 1µm × 1µm × 0.5µm water box as the target to calculate energy deposition on event-by-event basis and the three-dimensional coordinates of the interaction location, and then DBSCAN algorithm is used to calculate yields of SSB, DSB and cDSB in human fibroblast cell. The study investigated the influence of Linear Energy Transfer (LET) of protons, helium ions and carbon ions on the yields of DNA damages. Influence of cellular oxygenation on DNA damage patterns is investigated using MCDS code.Results. The study shows that DSB and SSB yields are influenced by the LET of the particles, with distinct trends observed for different particles. The cellular oxygenation is a key factor, with anoxic cells exhibiting reduced SSB and DSB yields, underscoring the intricate relationship between cellular oxygen levels and DNA damage. The study introduced DSB/SSB ratio as an informative metric for evaluating the severity of radiation-induced DNA damage, particularly in higher LET regions.Conclusions. The study highlights the importance of considering particle type, LET, and cellular oxygenation in assessing the biological effects of ionizing radiation.

Monte Carlo-based Investigation of Absorbed-dose Energy Dependence of Thermoluminescent Dosimeters in Therapeutic Proton and Carbon Ion Beams.

Background: The present study is aimed at calculating relative absorbed-dose energy response correction (R) of commonly used thermoluminescent dosimeters (TLDs) such as LiF, Li2B4O7, and Al2O3 as a function of depth in water for protons (50-250 MeV/n) and carbon ion (80-480 MeV/n) beams using Monte Carlo-based FLUKA code. Materials and Methods: On-axis depth-dose profiles in water are calculated for protons (50-250 MeV/n) and carbon ion (80-480 MeV/n) beams using FLUKA code. For the calculation of R, selective depths are chosen based on the depth-dose profiles. In the simulations, the TLDs of dimensions 1 mm × 1 mm × 1 mm are positioned at the flat, dose gradient, and Bragg peak regions of the depth-dose profile. Absorbed dose to detector was calculated within the TLD material. In the second step, TLD voxels were replaced by water voxel of similar dimension and absorbed dose to water was scored. Results: The study reveals that for both proton and carbon ion beams, the value of R differs from unity significantly at the Bragg peak position and is close to unity at the flat region for the investigated TLDs. The calculated R value is sensitive to depth in water, beam energy, type of ion beam, and type of TLD. Discussion: For accurate dosimetry of protons and carbon ion beams using TLDs, the response of the TLD should be corrected to account for its absorbed-dose energy dependence.

Evaluation of uncertainty in personal dose measured using CaSO4:Dy-based TLD badge at different workplaces.

The metrological quality of a measurement is characterised by evaluating the uncertainty in the measurement. In this paper, uncertainty in personal dose measured using individual monitoring CaSO4:Dy-based thermoluminescence dosimeter badge is evaluated by application of the guide to the expression of uncertainty in measurement method. The present dose reporting quantity, whole body dose (WBD) and the proposed quantity, personal dose equivalent, Hp(10) has been used as measurands. The influence of various input quantities on the measurement were analyzed through tests that conform to the requirements of the International Electrotechnical Commission IEC 62387. The study found that the expanded uncertainties for WBD and Hp(10) measurements were 63.4% and 41.4%, respectively, corresponding to a 95% coverage probability for workplace fields covering a wide photon energy range (33-1250 keV). However, the uncertainty estimates were quite lower for the type of workplaces that are identified using the dose evaluation algorithm. The input quantities, namely, the response to a mixture of photon beam qualities and photon energy and angular dependence contribute the most to the total uncertainty.

Monte Carlo-based dosimetry of proposed bi-radionuclide (125I and 106Ru/106Rh) eye plaque: A feasibility study.

BACKGROUND: Combining the sharp dose fall off feature of beta-emitting 106Ru/106Rh radionuclide with larger penetration depth feature of photon-emitting125I radionuclide in a bi-radionuclide plaque, prescribed dose to the tumor apex can be delivered while maintaining the tumor dose uniformity and sparing the organs at risk. The potential advantages of bi-radionuclide plaque could be of interest in context of ocular brachytherapy. PURPOSE: The aim of the study is to evaluate the dosimetric advantages of a proposed bi-radionuclide plaque for two different designs, consisting of indigenous 125I seeds and 106Ru/106Rh plaque, using Monte Carlo technique. The study also explores the influence of other commercial 125I seed models and presence or absence of silastic/acrylic seed carrier on the calculated dose distributions. The study further included the calculation of depth dose distributions for the bi-radionuclide eye plaque for which experimental data are available. METHODS: The proposed bi-radionuclide plaque consists of a 1.2-mm-thick silver (Ag) spherical shell with radius of curvature of 12.5 mm, 20 µm-thick-106Ru/106Rh encapsulated between 0.2 mm Ag disk, and a 0.1-mm-thick Ag window, and water-equivalent gel containing 12 symmetrically arranged 125I seeds. Two bi-radionuclide plaque models investigated in the present study are designated as Design I and Design II. In Design I, 125I seeds are placed on the top of the plaque, while in Design II 106Ru/106Rh source is positioned on the top of the plaque. In Monte Carlo calculations, the plaque is positioned in a spherical water phantom of 30 cm diameter. RESULTS: The proposed bi-radionuclide eye plaque demonstrated superior dose distributions as compared to 125I or 106Ru plaque for tumor thicknesses ranges from 5 to 10 mm. Amongst the designs, dose at a given voxel for Design I is higher as compared to the corresponding voxel dose for Design II. This difference is attributed to the higher degree of attenuation of 125I photons in Ag as compared to beta particles. Influence of different 125I seed models on the normalized lateral dose profiles of Design I (in the absence of carrier) is negligible and within 5% on the central axis depth dose distribution as compared to the corresponding values of the plaque that has indigenous 125I seeds. In the presence of a silastic/acrylic seed carrier, the normalized central axis dose distributions of Design I are smaller by 3%-12% as compared to the corresponding values in the absence of a seed carrier. For the published bi-radionuclide plaque model, good agreement is observed between the Monte Carlo-calculated and published measured depth dose distributions for clinically relevant depths. CONCLUSION: Regardless of the type of 125I seed model utilized and whether silastic/acrylic seed carrier is present or not, Design I bi-radionuclide plaque offers superior dose distributions in terms of tumor dose uniformity, rapid dose fall off and lesser dose to nearby critical organs at risk over the Design II plaque. This shows that Design I bi-radionuclide plaque could be a promising alternative to 125I plaque for treatment of tumor sizes in the range 5 to 10 mm.

Microdosimetry-based investigation of biological effectiveness of252Cf brachytherapy source: TOPAS Monte Carlo study.

Objective.To investigate biological effectiveness of252Cf brachytherapy source using Monte Carlo-calculated microdosimetric distributions.Approach.252Cf source capsule was placed at the center of the spherical water phantom and phase-space data were scored as a function of radial distance in water (R= 1-5 cm) using TOPAS Monte Carlo code. The phase-space data were used to calculate microdosimetric distributions at 1µm site size. Using these distributions, Relative Biological Effectiveness (RBE), mean quality factor (Q̅) and Oxygen Enhancement Ratio (OER) were calculated as a function ofR.Main results.The overall shapes of the microdosimetric distributions are comparable at all the radial distances in water. However, slight variation in the bin-wise yield is observed withR. RBE,Q̅and OER are insensitive to R over the range 1-5 cm. Microdosimetric kinetic model based RBE values are about 2.3 and 2.8 for HSG tumour cells and V79 cells, respectively, whereas biological weighting function-based RBE is about 2.8. ICRP 60 and ICRU 40 recommendation-basedQ̅values are about 14.5 and 16, respectively. Theory of dual radiation action based RBE is 11.4. The calculated value of OER is 1.6.Significance.This study demonstrates the relative insensitivity of RBE,Q̅and OER radially away from the252Cf source along the distances of 1-5 cm in water.

Monte Carlo Study on Dose Distributions Around 192Ir, 169Yb, and 125I Brachytherapy Sources Using EGSnrc-based egs_brachy User-code.

Introduction: As per the recommendations of the American Association of Physicists in Medicine Task Group 43, Monte Carlo (MC) investigators should reproduce previously published dose distributions whenever new features of the code are explored. The purpose of the present study is to benchmark the TG-43 dosimetric parameters calculated using the new MC user-code egs_brachy of EGSnrc code system for three different radionuclides 192Ir, 169Yb, and 125I which represent high-, intermediate-, and low-energy sources, respectively. Materials and Methods: Brachytherapy sources investigated in this study are high-dose rate (HDR) 192Ir VariSource (Model VS2000), 169Yb HDR (Model 4140), and 125I -low-dose-rate (LDR) (Model OcuProsta). The TG-43 dosimetric parameters such as air-kerma strength, S k, dose rate constant, Λ, radial dose function, g(r) and anisotropy function, F(r,θ) and two-dimensional (2D) absorbed dose rate data (along-away table) are calculated in a cylindrical water phantom of mass density 0.998 g/cm3 using the MC code egs_brachy. Dimensions of phantom considered for 192Ir VS2000 and 169Yb sources are 80 cm diameter ×80 cm height, whereas for 125I OcuProsta source, 30 cm diameter ×30 cm height cylindrical water phantom is considered for MC calculations. Results: The dosimetric parameters calculated using egs_brachy are compared against the values published in the literature. The calculated values of dose rate constants from this study agree with the published values within statistical uncertainties for all investigated sources. Good agreement is found between the egs_brachy calculated radial dose functions, g(r), anisotropy functions, and 2D dose rate data with the published values (within 2%) for the same phantom dimensions. For 192Ir VS2000 source, difference of about 28% is observed in g(r) value at 18 cm from the source which is due to differences in the phantom dimensions. Conclusion: The study validates TG-43 dose parameters calculated using egs_brachy for 192Ir, 169Yb, and 125I brachytherapy sources with the values published in the literature.

TISSUE-EQUIVALENCE OF H2 GAS FOR MICRODOSIMETRY IN NEUTRON FIELDS: A GEANT4 MONTE CARLO STUDY.

Hydrogen (H2) as filling gas in Tissue-equivalent proportional counter (TEPC) for measurements of microdosimetric distributions in neutron fields is investigated using the Monte Carlo-based Geant4 toolkit. The neutron fields considered are monoenergetic neutrons (1 keV-14 MeV) and ISO reference neutron sources 241Am-Be, 241Am-B, 252Cf and 252Cf + D2O. Based on the energy deposited in the gas cavities (tissue-equivalent (TE) propane and H2) of the TEPC, density scaling correction, ${f}_{\rho }$, to be applied on H2 gas to achieve tissue-equivalence is calculated for the above sources at 1µm site size. The calculated value of ${f}_{\rho }$ for the ISO-neutron sources is 0.40, which suggests that energy deposition in the gas cavities is predominantly due to crossers. In the case of monoenergetic neutrons, depending upon the energy, ${f}_{\rho }$ is in the range of 0.11-0.45. The study shows that the TE propane and H2-based microdosimetric distributions are comparable when the density of H2 is scaled appropriately. Mean quality factor $\overline{Q}$ calculated based on microdosimetric distribution increases initially with neutron energy up to 100 keV and thereafter decreases with energy. Depending upon the neutron energy, TE propane and H2-based $\overline{Q}$ values are comparable to within 3-15%, and for the ISO sources, the comparison is within 5-8%. For the ISO neutron sources, $\overline{Q}$ values are in the range of ~10-15.

Microdosimetry-based relative biological effectiveness calculations for radiotherapeutic electron beams: a FLUKA-based study.

Based on FLUKA, the present study is aimed at calculating the microdosimetric distributions of electron beams (6, 12 and 18 MeV) for radiotherapy as a function of depth in water at a site-size of 1 µm using a tissue-equivalent proportional counter (TEPC). Using the calculated microdosimetric distributions, the depth-specific relative biological effectiveness (RBE) of electron beams used in radiotherapy is calculated based on the theory of dual radiation action (TDRA) and the microdosimetric kinetic model (MKM). The TDRA-based calculation shows the variation of RBE of an electron beam with the absorbed dose and depth in water. In this study, we compared the RBE values calculated based on the TDRA and MKM. The FLUKA-based microdosimetric distributions in water obtained using the pre-calculated electron fluence spectra resulted in an improvement in the computational efficiency by a factor of 110 when compared with a full simulation. Depending on the beam energy and depth of water, RBETDRA was in the range 0.67-0.78. RBEMKM at 10% survival of HSG tumor cells was 0.84, which was nearly independent of the depth and beam energy.

Technical note: EGSnrc-based dosimetric study of the BEBIG 60Co HDR brachytherapy sources.

PURPOSE: The purpose of this work is to calculate two-dimensional (2D) dose rate distributions around the BEBIG (Eckert & Ziegler, BEBIG GmbH, Germany) models GK60M21 (old) and Co0.A86 (new) 60Co high dose rate brachytherapy sources in an unbounded liquid water phantom. The study includes calculation of absorbed dose to water-kerma ratio D/K around the BEBIG sources and a 60Co point source in water. A comparison is made with previously published data. METHODS: The EGSnrcMP Monte Carlo code system is used to calculate the absorbed dose and water-kerma in water and air-kerma strength in vacuum. EGSnrcMP-based user codes such as EDKnrc, FLURZnrc, and DOSRZnrc are employed in the work. RESULTS: The value of D/K reaches a maximum of 1.040 +/- 0.002 for the 60Co point source (constant between 3.6 and 4.5 mm from the source) and 1.076 +/- 0.002 for the BEBIG sources (constant between 2.6 and 3.2 mm along the transverse axis of the sources). Dose rate data for the new and old sources are comparable to published data for radial distances r > 0.5 cm. Differences up to 9% are observed at points close to the source (r = 0.25 cm). In addition for the new source, compared to previously published data, dose rate data are higher by 14% along the longitudinal axis where the source cable is connected. Dose rate differences on the longitudinal axis (8 = 180 degrees) of this source are explained by varying the length of the simulated source cable. CONCLUSIONS: The 2D rectangular data set calculated in the present work could be considered for quality control on radiotherapy treatment planning systems.

Monte Carlo investigation of energy response of various detector materials in ¹²5I and ¹69Yb brachytherapy dosimetry.

Relative absorbed-dose energy response correction R for different detector materials in water, PMMA and polystyrene phantoms are calculated using Monte Carlo-based EGSnrc code system for ¹²5I and ¹69Yb brachytherapy sources. The values of R obtained for ¹²5I source are 1.41, 0.92, 3.97, 0.47, 8.32 and 1.10, respectively, for detector materials LiF, Li2B4O7 , A12O3, diamond, silicon diode and air. These values are insensitive to source-to-detector distance and phantom material. For ¹69Yb source, R is sensitive to source-to-detector distance for detector materials other than air and Li2B4O7. For silicon, R increases from 3 to 4.23 when depth in water is increased from 0.5 cm to 15 cm. For ¹69Yb source, the values of R obtained for air and Li2B4O7 in PMMA and polystyrene phantoms are comparable to that obtained in water. However, LiF, Si and A12O3 show enhanced response and diamond shows decreased response in PMMA and polystyrene phantoms than in water.

APPLICABILITY OF PURE PROPANE GAS FOR MICRODOSIMETRY AT BRACHYTHERAPY ENERGIES: A FLUKA STUDY.

Applicability of pure propane gas for microdosimetric measurements at photon energies relevant in brachytherapy is studied using the Monte Carlo-based FLUKA code. Monoenergetic photons in the energy range of 20-1250 keV and brachytherapy sources such as 103Pd, 125I, 169Yb, 192Ir, 137Cs and 60Co are considered in the study. Using the calculated values of energy deposited in the sensitive region of LET-1/2 tissue-equivalent proportional counter filled with pure propane gas and tissue-equivalent propane gas, values of density scaling factor for the site sizes of 1 and 8 µm are obtained. The study shows that density of propane gas should be lowered by a factor of about 0.93 for 169Yb, 192Ir, 137Cs and 60Co sources for the site sizes of 1-8 µm. For 125I source, the density of propane gas requires a scaling of 0.93 for 1 µm site size, whereas for site sizes 2-8 µm, density need not be altered. 103Pd source does not require density scaling for site sizes 1-8 µm.

EGSnrc-based depth-dependent photon energy response and phantom scatter corrections for low-energy brachytherapy sources.

In the present study, beam quality correction, [Formula: see text], and phantom scatter correction, kphan(r), for low-energy brachytherapy sources, 131Cs, 125I, and 103Pd, are calculated using the Monte Carlo-based EGSnrc code system as a function of the distance along the transverse axis of the source. The solid-state detectors investigated are diamond, LiF, Li2B4O7, Al2O3, and radiochromic films, such as HS, EBT, EBT2, EBT3, RTQA, XRT, and XRQA. The solid phantoms investigated are polystyrene, PMMA, virtual water, solid water, plastic water (LR), A150, RW1, RW3, and WE210. For a given detector and brachytherapy source, [Formula: see text] is independent of distance in the water phantom. Meanwhile, for a given detector, kphan(r) depends on the distance from the source for the investigated solid phantoms. Moreover, the kphan(r) values do not change with the detector type for sources 131Cs, 125I, and 103Pd at all distances. The LR and A150 phantoms are water equivalent for the investigated distances of 1-5 cm. The phantoms including solid water, virtual water, and WE210 are not water-equivalent for distances beyond 1 cm. Furthermore, PMMA, polystyrene, RW1, and RW3 are not water equivalent.

MACHINE LEARNING ALGORITHMS FOR IDENTIFICATION OF ABNORMAL GLOW CURVES AND ASSOCIATED ABNORMALITY IN CaSO4:DY-BASED PERSONNEL MONITORING DOSIMETERS.

In the present study, machine learning (ML) methods for the identification of abnormal glow curves (GC) of CaSO4:Dy-based thermoluminescence dosimeters in individual monitoring are presented. The classifier algorithms, random forest (RF), artificial neural network (ANN) and support vector machine (SVM) are employed for identifying not only the abnormal glow curve but also the type of abnormality. For the first time, the simplest and computationally efficient algorithm based on RF is presented for GC classifications. About 4000 GCs are used for the training and validation of ML algorithms. The performance of all algorithms is compared by using various parameters. Results show a fairly good accuracy of 99.05% for the classification of GCs by RF algorithm. Whereas 96.7% and 96.1% accuracy is achieved using ANN and SVM, respectively. The RF-based classifier is recommended for GC classification as well as in assisting the fault determination of the TLD reader system.