A team of researchers in Berlin saw that high costs and a complex manufacturing process were limiting the use of anthropomorphic phantoms of patients in radiology. There must be more affordable and efficient way, they thought.

And it turns out they were right.

Lead author Michael Scheel, MD, and colleagues from the department of radiology at Charité-Universitätsmedizin Berlin looked to develop a better method for creating anthropomorphic phantoms, sharing their findings in Radiology.

Scheel et al. noted that numerous attempts have been made in the past to develop affordable phantoms, but it has proven to be a problematic endeavor.

“Despite the use of advanced technologies, these methods remain complex and rely on the principle of combining a limited number of different materials and Hounsfield units,” the authors wrote. “Therefore, the resulting phantoms cannot reproduce the full range of Hounsfield units and anatomic heterogeneity that is present within the human body.”

What they did

The researchers pushed ahead with the idea that they could take advantage of a technology being done more and more in modern scientific research: 3D printing.

“Ideally, a method to produce realistic anthropomorphic phantoms of individual patients should allow for precise distribution of radiopaque material in three dimensions and be simple and cost-effective,” the authors wrote. “In this context, conventional inkjet technology in combination with modified, iodine-enhanced ink and stacking of printed images presented an opportunity to create three-dimensional objects with desired shape and CT attenuation. Furthermore, gray-scale-encoded CT images may be used to directly transfer anatomic details and attenuation characteristics of a patient into a three-dimensional object.”

The team used a standard inkjet printer and replaced the inkjet cartridges with “an aqueous potassium iodide solution of 600 mg/ml.” Prints were made on plain paper and that paper was stacked in a box made of medium-density fiberboard.

Starting with an upper-abdomen CT scan of a 25-year-old male patient, they printed 10 consecutive sections of the scan, resulting in a stack of 100 pieces of paper. The stack was scanned by a CT scanner, and it showed “a surprisingly detailed reproduction" of the original scan.

“The results show that a conventional inkjet printer can be used in combination with modified, iodine-enhanced ink to produce three-dimensional objects with the shape and attenuation characteristics of patients,” the authors wrote. “Production of the phantoms was simple and fast, yet accurate and reliable. Results indicate that the method may be applied to any desired forms or patient templates.”

There are numerous uses for this method in radiology, Scheel and colleagues said, including “protocol optimizations, device testing, calibration, quality assurance, dosimetric investigations, staff training, and demonstration purposes.”

The team’s research did have limitations, they explained. For instance, the Hounsfield units of the phantom “did not exactly match” the actual patient’s scan, though it was certainly close. Also, the method was tested on one final patient phantom, so more research is needed to confirm the validity of the team’s results.