Assessing Students' Anatomical Knowledge on Bones, Commercial Models, and 3D Prints

3D prints have been increasingly used as substitutes for human tissue within medical and educational settings. Several studies demonstrated that anatomical structures could be 3D printed with the accuracy needed in anatomical education. These studies, however, focused mainly on the shape and did not take into account finer features such as texture and colour. This study aimed to investigate students' test performance on real bones, commercial anatomical models and 3D prints produced on a desktop printer. A total of 211 students (divided into three groups) in a musculoskeletal anatomy course were asked to complete a practical test on vertebral anatomy. In the test, at each of the nine stations, a vertebra (real, model or 3D printed) was presented and students were asked to identify which vertebral region it belonged to, and a specific anatomical structure. The sequence of real, model and 3D printed bones presented along the stations was different for each group, to control for possible order effects. There were no significant differences in identifying vertebral regions or larger structures, such as transverse and spinous processes, across the three types of bones. However, significant difference was found in the identification of smaller structures, such as epiphyseal rim (p < 0.0005) and demifacets of thoracic vertebra (p < 0.05), with the highest percentage of correct answers for real bones, followed by 3D prints. The results suggest that students recognise anatomical structures equally well on real, model or 3D printed bones and that 3D prints are better than models for identifying smaller structures. This supports the view that some anatomical structures can be 3D printed with the accuracy required in anatomy education, even if produced on desktop printers.