Automated tooth segmentation as an innovative tool to assess 3D-tooth movement and root resorption in rodents

Background Orthodontic root resorptions are frequently investigated in small animals, and micro-computed tomography (μCT) enables volumetric comparison. Despite, due to overlapping histograms from dentine and bone, accurate quantification of root resorption is challenging. The present study aims at (i) validating a novel automated approach for tooth segmentation (ATS), (ii) to indicate that matching of contralateral teeth is eligible to assess orthodontic tooth movement (OTM) and root resorption (RR), (iii) and to apply the novel approach in an animal trial performing orthodontic tooth movement. Methods The oral apparatus of three female mice were scanned with a μCT. The first molars of each jaw and animal were segmented using ATS (test) and manually (control), and contralateral volumes were compared. Agreement in root volumes and time efficiency were assessed for method validation. In another n = 14 animals, the left first upper molar was protracted for 11 days at 0.5 N, whereas the contralateral molar served as control. Following ATS, OTM and RR were estimated. Results ATS was significantly more time efficient compared to the manual approach (81% faster, P < 0.01), accurate (volume differences: − 0.01 ± 0.04 mm3), and contralateral roots had comparable volumes. Protracted molars had significantly lower root volumes (P = 0.03), whereas the amount of OTM failed to reveal linear association with RR (P > 0.05). Conclusions Within the limits of the study, it was demonstrated that the combination of ATS and registration of contralateral jaws enables measurements of OTS and associated RR in μCT scans.