Objectives To evaluate the surface and positional trueness of removable dies fabricated using fused filament fabrication (FFF), compared to dies produced with digital light processing (DLP), addressing the limited evidence on the deviations of FFF-based removable dies.Methods A typodont with a prepared right mandibular first molar was digitized to design a removable die and corresponding hollow partial arch cast. Forty dies (n = 10) and four casts (n = 1) were fabricated using FFF (Filadent Aligner FF-AL, Filadent Gypsum FF-GP, Filadent Tray FF-TR) or DLP (DentaModel DM). Dies and their positions in the casts were digitized to assess surface (crown, root, root base, overall) and positional (crown region and point-based) deviations. Data were analyzed either with one-way analysis of variance (surface deviations) or Kruskal-Wallis tests (positional deviations, α = 0.05).Results DM exhibited the lowest crown region and overall deviations, followed by FF-TR (P < 0.001). Root deviations increased in the order of DM, FF-TR, FF-AL, and FF-GP, whereas FF-TR had the lowest base of the root deviations (P < 0.001). Seated FF-GP dies had lower crown region deviations than the other FFF-based dies, while DM led to lower deviations than FF-TR (P ≤ 0.041). FF-AL showed lower point-based deviations than FF-GP (P = 0.001).Conclusions FFF-fabricated dies showed lower surface trueness than DM dies, with FF-TR achieving the highest trueness among FFF dies. Positional deviations remained within clinically acceptable limits, though FF-TR dies tended to be positioned coronally and the others apically. FFF may be used to produce clinically acceptable dies, but DLP offers superior surface trueness.Clinical Significance Removable dies produced with the tested polylactic acid filaments and FFF printer may serve as alternatives to those fabricated with the tested resin and vat polymerization printer, given the small and potentially clinically irrelevant differences in surface deviations and clinically acceptable mean positional deviations.
Çakmak et al. (Wed,) studied this question.