Comparative Analysis of Additive Manufacturing Methods to fabricate Auricular Prostheses

Abstract

In the current literature to the topic only little was published on the geometrical accuracy and resemblance of AM-produced prostheses replicas. Therefore the main objective of the present study was to identify the superior AM method from FDM, SLS and SL in terms of dimensional accuracy, skin details reproduction and efficiency to produce APRs using thereby the rapid prototyping approach. Twenty three subjects underwent a clinical study procedure encompassing ear anthropometry, followed by structured light scanning of patients’ left auricles. The auricular area including the pinna was scanned with a portable surface scanner (Artec 3D Spider, Artec Group, Luxembourg, Luxembourg), utilizing the structured light scanning method. The distances of the anthropometrical landmarks were measured within the Software (ARTEC Studio, version 9), three times blinded with the “digital lineal”. After the gathered data was post-processed and converted into OBJ format, it has been transferred to AM machines to produce 57 APRs by means of FDM (n=23), SLS (n=23) and SL (n=11) methods. The manufactured APRs were measured blinded three times each distance between the landmarks with the digital calipers. Measurements gathered from APRs have been compared to the In-vivo and CAD data groups. Results have been statistically evaluated. Additionally, the surface analysis of APRs utilizing stereomicroscopy and profilometry was conducted to ascertain what level of skin details reproduction is achievable. Production costs and time were calculated. The analysis of dimensional accuracy revealed difference up to 0.56 mm. This was found clinically acceptable, as not exceeding the threshold of 2 mm (see on page 20), which was set as a threshold ( However, the comparison of relative mean differences disclosed the bias of up to 1.85 % between the in vivo data group and AM-produced APRs, which was higher than 1.5 %, as assumed in the present study. The comparison of relative mean differences between CAD data group and APRs did not reveal any discrepancies that may be clinically recognizable. As far as pure accuracy of AM methods is concerned, the FDM showed the best result. The reproduction of skin surface structure was only feasible where the skin details exceed 192 µm of depth. The reference wrinkles “Lobula basis” and “Lobula corpus” were visible on each APR. However, the wrinkle “Helix” was not reproducible by any of the employed AM methods. The FDM showed the most detailed reproduction of the tissue portion captured by means of structured light scanning. The staircase effect remains the main limiting factor of this AM method. The disclosed differences were found to be clinically acceptable, although in 5 of 42 comparisons the mean relative differences between in vivo and APRs exceeded slightly the threshold of clinical relevance set in the present study. The step of digital data acquisition was obviously more responsible for the revealed dimensional errors than the AM methods themselves. The method of FDM showed the best trade-off between dimensional accuracy, level of texture details and pricing. Thus FDM can be recommended for rapid and efficient manufacturing of prostheses replicas.