3D Printing: Part 2, radiation imaging and therapy applications

Introduction

Radiotherapy phantoms consist of tissue-equivalent materials formed as a simple geometry cast. They can also be made more sophisticated to emulate human-like anatomical geometries. But current state-of-art phantoms are constrained. They are limited to a bulk average for the physical anatomy and pathology of a healthy person. To overcome this limitation and to make phantoms that represent the anatomical and physical variations of a specific patient, the technique of 3D printing (also known as Additive Manufacturing (AM)), described in Part 1, is one of the most rapidly emerging and fast adapted technologies for affordable manufacturing of customised phantoms for radiation therapy.

QA of radiation therapy planning

The X-ray CT simulator planning procedure

The first step in the radiation therapy planning procedure involves obtaining computer tomography (CT) images of the patient. For the CT X-ray images, the patient lies on the CT scan examination couch the same way (complete with any immobilisation devices that hold the patient still) as when treated on the linear accelerator. It’s referred to as a ‘CT simulation’.

The patient automatically moves through the CT simulator turret for the examination and the X-ray transmission exposures are measured and stored as a series of 2D images. The 2D image data obtained throughout the length of the patient’s exposure can then be converted by computer calculations to a 3D matrix of points with values expressed as CT numbers or Hounsfield units (HU).

The HU for every point in the 3D matrix is then converted to electron densities for the 3D dose calculation by the computer treatment planning system (TPS). See section below for more detail on the procedures for HU versus density calibration.

The planning and radiation therapy workflow

Before any new treatment procedure begins, the dose calculated by the TPS must be confirmed by phantom measurements taken on the treatment machine. The measured phantom dose is compared against the planning computer calculated dose. Once the QA dose check shows a consistent accuracy per regulatory standards, the new technique is approved and becomes a routine treatment procedure (see Figure 1 for the overall workflow).

Anthropomorphic phantoms

The need for cheap customised phantoms

For patients treated by the standard fixed beam technique, measurements can be obtained using (i) a fixed block made of water equivalent material (for 1 or 2-dimensional measurements in a phantom of simulated normal tissue) or (ii) the commercial anthropomorphic phantom with tissue-equivalent materials that emulate the megavoltage X-ray physical characteristics (i.e. X-ray Photoelectric and Compton absorption) for soft tissue, bone and lung tissue.

Commercial anthropomorphic phantoms are sophisticated engineering and science-based products that emulate human-like tissue-equivalence and normal anatomical shapes. They are specifically designed for radiotherapy QA of routine treatment planning procedures. These phantoms are relatively high in cost due to the specialised moulding and casting of the materials and the manufacturing processes. They  lack flexibility for simulating customised dose checks needed for specific patients.

Routine use of anthropomorphic phantoms

Commercial anthropomorphic phantoms are designed to have average anatomical shapes and tissue-attenuations of a healthy living person. The phantom is made up of sectional slices  that are fabricated by special moulding and casting processes and mimic geometry and densities of lung tissue, soft tissue, and bone — using tissue-substitute materials that possess the same elemental composition and/or proportion (by weight) to human tissues, making them highly reliable and durable (see Figure 2).

However, the commercial phantoms have a number of limitations. They are:

• high manufacturing and material costs.
• difficult to customise (e.g. some phantoms only allow the insertion of generic tumour shapes and sizes);
• inaccurate representation of individual patients receiving treatment. Variations in organ physical and pathological detail can cause unavoidable uncertainties in the dose measured; and
• comprehensive end-to-end testing of complex treatment plan calculations are particularly impractical for obese patients or patients who have tiny or cavitating lesions.

Furthermore, as dynamic x-ray intensity modulated radiation therapy (IMRT)) became a routine clinical technique during the 2000-2010 decade, tools to measure the accuracy and consistency of the dose distribution, were limited. The solid block and commercial anthropomorphic phantom could not be easily used to check the accuracy of computer-generated 3D dose treatment plans for IMRT cases (see Figure 3).

The application of 3D Printing for radiation therapy

Disruptive 3D printing technology has been used for a range of innovative surgical, tissue engineering and other medical healthcare applications. It’s particularly useful for making individual customised 3D printed parts which a surgeon can use as: anatomical models for treatment planning or training procedures; implants when operating on a patient; or tools/guides to aid patient surgical procedures.

Over the years, due to its low manufacturing costs and its ability to customise tools for specific applications, 3D printing has provided opportunities for radiation therapy researchers to explore its capabilities and advantages for a range of imaging and dosimetry tasks. Their research and clinical interests were focused on the design and fabrication of customised phantoms that verify the radiation therapy treatment planning dosimetry calculations [1-4]. The availability of 3D printed devices for verifying patient dosimetry, became a fourth research level of interest (Figure 4).

3D printed devices or moulds can improve the accuracy of dose received by radiation therapy patients. Some 3D printed device examples are:

• immobilising and positioning devices that restrain patient movement during treatment (see Crowe article);
• compensators and bolus devices used to improve the patient’s dose distribution received due to irregular surfaces (see Crowe article);
• catheter treatment guides inserted in brachytherapy moulds; and
• anthropomorphic type phantoms used for imaging or dosimetric tasks (Figure 4).

Clinical workflow for using 3D printed phantoms

Figure 5 shows a standard workflow for the fabrication of AM anthropomorphic phantoms.

The steps are:

1. Diagnostic images of the patient’s CT scan) are acquired.

• 3D models of the patient’s tissue/organ segments are generated by either manual or automatic software processes using the cross-sectional CT scan data. A popular software package used for this process, is 3D Slicer.

• The data file of the 3D generated models are then converted to a GCODE file. This has a universal language for material extrusion and describes x, y, z movement of the printer.

GCODE also defines print speed, temperature, infill pattern and percentage parameters.

And lastly,

• Choose the printing process.

For Material Extrusion, the 3D printed phantoms can be:

• shells filled with tissue-substitute materials; or
• solid with different 3D print materials, infill patterns or infill per cent density.

Customised materials

Materials that aim to emulate human tissues and are used for X-ray radiation dosimetry are termed as ‘tissue-substitutes’ [5]. They can further be categorised as:

1.   Tissue-equivalent – materials that are not only identical in elemental constituents, but also identical in proportion by weight to human tissues;

or,

2.    Quasi-equivalent – materials that have the same elemental composition as human tissues except for Carbon (C) and Oxygen (O). In that case, the sum (C + O) is identical to human tissue organs/structures.

More advanced phantoms use real human skeleton such as the Erler Zimmer Natural Bone Full Body X-Ray Phantom – 7200 and QUART X-ray anthropomorphic phantoms.

To assess whether a 3D print material suitably emulates human tissues, the material must satisfy the same physical characteristics as X-ray absorption in tissue:

            (i) Photoelectric – indicated by CT Hounsfield Units;

and,

(ii) Compton attenuations – indicated by graphing CT Hounsfield Units against electron densities and elemental composition.

Dosimetry tests for diagnostic imaging

Photoelectric attenuation is a dominant effect at x-ray energies less than 200 keV (e.g imaging kilovoltage levels used for CT, MRI and PET scans).

For dosimetry measurements that simulate X-rays for diagnostic imaging, the 3D print materials must have the same Hounsfield Units as the emulated human anatomy.

Dosimetry tests for megavoltage X-ray therapy

Compton attenuation is the dominant x-ray absorption process (proportional to dose) for X-ray energies from 200 keV to 10MeV. The dose absorbed depends on the effective electron density.

For dosimetry measurements that simulate megavoltage X-ray therapy, the 3D print materials must have the same electron densities as the emulated human anatomy.

In view of this criteria, commercial 3D print materials are limited in how they can emulate human tissues (such as lung, soft tissue and bone) for dosimetry measurements at the diagnostic and therapy X-ray energies.

Calibration of 3D print materials

Cylindrical samples of 3D print materials and commercial samples of known electron density were together inserted in a calibrated HU phantom (Figure 6 (a)). The sample loaded phantom was CT scanned according to the standard CT protocol of 140 kVp and slice thickness increments of 2 mm.

CT image data was then transferred to the Varian Eclipse (V15.6) treatment planning system. For each insert sample, a volumetric region of interest was selected and contoured to measure mean HU and its standard deviation (±SD). The results were plotted against relative electron density (ρ_e/w) _­obtained from a Philip Big Bore CT calibration curve (Figure 6 (b)).

Optimising print materials for dosimetry

Unfortunately, a suitable 3D print material that emulates both the elemental composition and proportions (weight) of human tissues, is not yet available. A fundamental challenge at this stage is to develop materials with current state-of-the-art 3D print technologies that better emulate human tissues and organ structures.

It’s well known that elemental composition of print materials strongly influences their physical properties (i.e. density, melting point, thermal conductivity, electrical conductivity (resistivity), thermal expansion, corrosion resistance, etc.).

Even if the materials could be manufactured in the method of 3D printing, the extra specialised processes, outsourcing of raw materials, the materials themselves and manufacturing would make total costs significantly higher.

Custom-made phantoms

Although 3D printing has historically been employed for many years in a wide range of fields, optimal utilisation by researchers working in the medical radiation field of diagnostic and therapy, there’s still much research to successfully complete.

The fundamental question raised is:

How can we utilise 3D print technology to make fully functional, low-cost, and customisable phantoms suitable for diagnostic and therapy radiation dosimetry research and QA work?

Numerous innovations to satisfactorily address this question by medical physics and biomedical engineering research and development are underway.

It’s now possible to make affordable, patient-specific phantoms for imaging or dosimetry research and QA tasks using commercially available 3D printers and materials (Figure 7).

Figure 7 shows a 3D printed phantom designed for radiation therapy dosimetry tests. It was used to test the dosimetry of the CT scan image and the planning computer 3D dose calculations. Dose films are inserted in the 3D printed phantom to measure and check the TPS calculated dose and the treatment machine dose which the patient would ordinarily receive during a daily course of treatment.

References

[1] Tino, R., Leary, M., Yeo, A., Kyriakou, E., Kron, T., & Brandt, M. (2020). Additive manufacturing in radiation oncology: a review of clinical practice, emerging trends and research opportunities. International Journal of Extreme Manufacturing, 2(1), 012003.

[2] Tino, R., Yeo, A., Leary, M., Brandt, M., & Kron, T. (2019). A systematic review on 3D-printed imaging and dosimetry phantoms in radiation therapy. Technology in cancer research & treatment, 18, 1533033819870208.

[3] Rooney, M. K., Rosenberg, D. M., Braunstein, S., Cunha, A., Damato, A. L., Ehler, E., … & Golden, D. W. (2020). Three‐dimensional printing in radiation oncology: A systematic review of the literature. Journal of applied clinical medical physics, 21(8), 15-26.

[4] Filippou, V., & Tsoumpas, C. (2018). Recent advances on the development of phantoms using 3D printing for imaging with CT, MRI, PET, SPECT, and ultrasound. Medical physics, 45(9), e740-e760.

[5] White, D. R., Martin, R. J., & Darlison, R. (1977). Epoxy resin based tissue substitutes. The British journal of radiology, 50(599), 814-821.

To be continued:

Part 3. of this series will provide a more in-depth detail on these custom-made phantoms and the MEX 3D print techniques and materials that enable us to make phantoms that emulate the CT scan HU values of lung tissue, soft tissue, and bone densities.

Rance Tino PhD, 2 April 2022

Help support mainstream medical physicists and biomedical engineers in R&D and collaborative multidisciplinary projects……..►►►Donate Now