INTRODUCTION

health canada’s Human Monitoring Laboratory (HML), which operates the Canadian National Calibration Reference Centre for In Vivo Monitoring (Kramer and Limson Zamora 1994), has previously conceptualized and described a sliced planar lung set for the Lawrence Livermore National Laboratory (LLNL) phantom (Kramer and Hauck 2000). The HML uses lung sets based on this model for its performance testing program of lung counting systems (Kramer et al. 1997). The performance testing program is part of the Canadian Nuclear Safety Commission’s regulatory requirements as set out in document REGDOC 2.7.2, Dosimetry, Vol 2: Technical and management system requirements for dosimetry services (CNSC 2012).

The original blank lung set was obtained from Radiology Support Devices (RSD, Inc., Long Beach, CA) and is made of low-density polyurethane foam in order to mimic lung tissue density. This set was cut into 2-cm sections in the transverse direction by RSD to allow placement of sealed, radioactive planar lung inserts.

These planar inserts are prepared using HCl, which can contribute to the degradation of the laminated filter paper over time and can increase the risk of contaminating the lung set. Decontamination of the RSD blank lung set can be difficult due to its porosity. The decontamination process requires soaking each of the foam slices in a decontamination solution multiple times and leaving them to air dry. Then, the blank set must be counted to ensure residual contamination is eliminated.

As a result of decontamination and general wear-and-tear, the HML has witnessed the degradation of the foam lung set over time. This has motivated the HML to develop a new method to produce blank lung sets that are more durable, easier to clean, and/or less expensive to replace.

The proposed solution was to 3D-print a replica lung set that would match the RSD foam set in important aspects such as size, shape, and density. Although elemental composition is important in radiation measurements (Kramer 1996), detailed elemental composition information on the 3D-printing filament (thermoplastic polyurethane, TPU) was not available and not given by the manufacturers of the materials. Previous research on the topic (Changfang et al. 2017) led to general information but no specifics; however, the measurements conducted during this validation exercise have shown that it is a suitable tissue-equivalent material for creating 3D-printed lung set used for the calibration of lung counting system.

Another advantage of using 3D-printing technology is its increased availability and affordability, as well as the growing amount of information on the topic available online. The development of the 3D lung set model by the HML allows sharing the manufacturing information and instructions easily with other lung counting facilities, which supports further improvements in radiation protection.

This paper describes the construction of 3D-printed blank lung sets and compares efficiencies obtained between the two sets using spiked planar inserts in a lung-counting system.

MATERIALS AND METHODS Creating the 3D-printed lung sets The original blank foam set

The original blank foam set was constructed by RSD from low-density polyurethane foam and was made to fit the second-generation LLNL chest phantom (Fig. 1). The blank set contains no radioactivity. The set was custom sliced transversely by RSD (RSD, Inc., Long Beach, CA) into 2-cm increments to allow the placement of HML’s planar lung inserts (Kramer and Hauck, 2000). Fig. 2 is showing the original RSD foam set beside the new 3D-printed version produced at the HML.

Fig. 1:

Lung foam sets of the HML phantom, made to fit the second generation of the Lawrence Livermore National Laboratory (LLNL) chest phantom.

Fig. 2:

Original RSD lung set (left) and its 3D-printed version.

Development of a new 3D-printed set

For the development of a 3D-printed lung set, the first step was to digitize a model of the 2-cm lung slices from the blank foam set. This was accomplished by scanning each individual lung slice using a laser digitizer (Makerbot 3D Desktop Digitizer, model number MP03995; Makerbot Industries, New York, NY). The black foam slices were covered with talcum powder to improve light reflection, allowing the scanner to better resolve the image.

The second step was to import the resulting scanned files into a MakerBot Replicator 2X 3D printer (Makrbot Industries, New York, NY). This allowed us to experiment with different printer settings, such as the infill density (fullness of the inside of the print), wall thickness size, and layer height, in order to reproduce each piece with dimensions and weight as close as possible to the original foam piece. Each lung slice was individually printed using a flexible filament (Ninjaflex 85A flexible polyurethane 1.75 mm diameter filament; Ninjaflex, Lititz, PA).

Table 1 presents the specifications, including the volume, weight, and the density of the two kinds of lung sets (RSD or 3D-printed at HML). Table 2 shows that the weight of the 3D-printed lung set and the weight of the original RSD foam match closely, with an overall difference of 5.6%.

Table 1 - Lung set specifications. Lung Set Sides Volume Weight Density (cm3) (g) (g cm−3) RSD foam left 1669.58 405 0.2426 right 2220.85 580 0.2612 HML 3D-printed left 1669.58 410 0.2456 (infill 20%) right 2220.85 520 0.2341 HML 3D printed left 1669.58 495 0.2965 (infill 25%) right 2220.85 665 0.2994
Table 2 - Weights of lung blank sets: RSD foam set vs. 3D-printed set. Piece(s) RSD foam set 3D-printed set Difference L = left R = right weight (g) weight (g) % L1 18.91 15.52 −17.9 L2 37.74 36.43 −3.5 L3 52.4 55.29 5.5 L4 64.57 66.83 3.5 L5 65.31 67.22 2.9 L6 61.12 57.22 −6.4 L7 51.15 49.51 −3.2 L8 42.12 42.96 2.0 L9 12.21 14.387 17.8 R1 26.1 21.93 −16.0 R2 36.14 31.21 −13.6 R3 54.38 46.99 −13.6 R4 67.23 62.36 −7.2 R5 80.03 70.47 −11.9 R6 84.66 76.2 −10.0 R7 87.77 80.92 −7.8 R8 91.05 79.11 −13.1 R9 51.49 54.91 6.6 Total mass 984.4 929.5 −5.6
Planar lung inserts

The HML routinely pairs planar lung inserts with the RSD blank lung set for its performance testing program of lung-counting systems. Planar lung inserts (Fig. 3) consist of 16 double-laminated filter paper pieces (eight pieces each for the left and right lungs). These inserts are placed in between the blank lung slices and have been shown to reproduce the effect of a homogenous deposition of radioactivity in the lungs (Kramer et al. 1999). In this validation exercise, natural uranium (Nat U), 241Am, and 152Eu inserts were used to test the performance of the 3D-printed lung set vs. the RSD foam set.

Fig. 3:

Planar inserts used as sources inside lung sets.

EXPERIMENTATION

The RSD foam and 3D-printed lung sets were loaded with planar inserts and counted in the HML lung-counting system. The system is composed of four P-type germanium detectors with a crystal dimension of 85 mm in diameter and 30 mm in thickness. The inactive layer is 700 μm thick on the surface. The entrance window is 0.76 mm carbon fiber and is 5 mm from the germanium crystal.

RESULTS AND DISCUSSION

Measurements at various energies were completed using planar lung inserts containing americium, uranium, and europium sources. The same counting time and geometry were used to obtain the efficiencies and calculated bias.

The bias was calculated using the following equation:

Bias%=EfficiencyRSD−Efficiency3DEfficiencyRSDx100.

The detector efficiencies obtained using the RSD foam set as well as the 3D-printed set were compared and are presented in Table 3.

Table 3 - Total efficiency (E) and Bias obtained using the RSD vs. 3D-printed lung sets, measuring 241Am, 152Eu, and natural uranium lung inserts. Energy Isotope RSD foam set 3D-printed set Bias (KeV) E(count/photon) E(count/photon) % 17.5 241Am 2.97 × 10−4 3.61 × 10−4 22 26.6 241Am 5.08 × 10−3 6.30 × 10−3 24 59.54 241Am 1.59 × 10−2 1.61 × 10−2 1 63.3 Nat U 1.57 × 10−2 1.8 × 10−2 15 92.6 Nat U 1.36 × 10−2 1.53 × 10−2 12 143.76 Nat U 1.82 × 10−2 2.04 × 10−2 12 185.72 Nat U 1.69 × 10−2 1.83 × 10−2 9 40.12 152Eu 2.17 × 10−2 2.22 × 10−2 2 45.40 152Eu 1.63 × 10−2 1.67 × 10−2 2 121.78 152Eu 2.02 × 10−2 1.97 × 10−2 −3 244.67 152Eu 1.43 × 10−2 1.42 × 10−2 −1 344.30 152Eu 1.17 × 10−2 1.15 × 10−2 −2 411.09 152Eu 8.97 × 10−3 8.63 × 10−3 −4 444 152Eu 1.00 × 10−2 9.78 × 10−3 −2

The results for the efficiencies show an overestimation of the radioactivity by 22% and 24% at 17 keV and 26 keV, respectively, and a maximum bias of 15% for photon energies between 60 keV and 444 keV. If compared with other parameters already studied by the HML (lung volume, respiratory motion, deposition patterns, etc.) (Kramer et al., 2004), these values show that the bias obtained using the new 3D-printed inserts will have a minimal contribution to the overall uncertainty in lung counting and can be neglected.

CONCLUSION

The HML has used the RSD blank foam lung set from its performance testing program as a model to create a 3D-printed version. The use of a 3D-printing scanner allowed the production of an identically shaped piece for each lung slice, producing a blank lung set that can be cheaply and easily created using a standard desktop 3D printer. When compared to each other, the RSD foam set and the new 3D-printed set differed in weight by 5.6% overall.

The biases obtained when comparing the two different lung sets using planar inserts containing natural uranium (Nat U), 241Am, and 152Eu were all below 15% at energies above 40 KeV. The HML considers these results acceptable when compared to other known errors that can affect lung-counting results (such as positioning errors), and considering the advantages obtained from the 3D-printed lung set.

These results show that the new 3D-printed lung set is a suitable replacement for the old RSD lung set and that this change does not have a significant influence on the counting efficiencies obtained using the HML lung counting system.

REFERENCES Canadian Nuclear Safety Commission. REGDOC 2.7.2, Dosimetry, vol 2: technical and management system requirements for dosimetry services. 2012. Changfang G, Priyadarshini L, Cormier D, Pan L, Tuber J. A preliminary study of cushion properties of a 3D printed thermoplastic polyurethane Kelvin foam. Packaging Technol Sci 31(3):1–8; 2017. Kramer GH, Limson Zamora M. The Canadian National Calibration Reference Centre for Bioassay and In Vivo Monitoring: a program summary. Health Phys 67:192–196; 1994. Kramer GH, Hauck B. The Human Monitoring Laboratory, Radiation Protection Bureau: sliced lung sets, an alternative to whole lung sets? Radiat Protect Dosim 89:309–312; 2000. Kramer GH, Hauck B, Lee TY, Chang SY. The Human Monitoring Laboratory, Radiation Protection Bureau: comparison of sliced and whole lung sets for the LLNL and JAERI torso phantoms using Ge detectors. Health Phys 76:547–552; 1999. Kramer GH, Hauck BM, Marro L, Inn KGW, Unterweger M, Hodge P. The use of autoradiography for investigating the distribution of radioactivity in lung counter calibration sources. Health Phys 86:359–364; 2004. Kramer GH, Burns LC, Yiu S. Lung counting: evaluation of uncertainties in lung burden estimation arising from a heterogeneous lung deposition using Monte Carlo code simulations. Rad Protect Dosim 74(3):173–182; 1997.