Optical Interocclusal Records are Repeatable and Accurate

Methods: A jig was created which held a maxillary dental model rigidly, and allowed precise movement of the mandibular model, using an XYZ positioning table. A 3D optical scanner was constructed using off-the-shelf components, capable of recording all upper and lower incisors/canines simultaneously. The mandibular model was positioned to simulate left lateral excursion and 5 interocclusal scans were taken.  This was repeated for right, protrusive, retrusive, open and closed positions.  The stone models were then digitized (Rexcan DS2,Europac 3D,Crewe) and 3 reference points were identified on both virtual models (LL6,LR6,UR6,UL6,LL1 and UR1). Custom software registered the scanned models with each optical interocclusal scan in turn, and the positions of the mandibular reference points, relative to the maxillary reference points were assessed. Repeatability was defined as the standard deviation from the mean for each set of 5 scans.

Next, the mandibular model was positioned near maximum intercuspation. An interocclusal scan was taken then the mandibular model was lowered by 100µm. This was repeated 10 times until the model had been lowered by 1mm. The virtual models were registered to these scans, and the calculated distances traversed by the mandibular reference points between each scan were compared to the known distances.

Results:  The repeatability (pooled standard deviations) of the derived positions of the mandibular reference points were 22µm(LL1), 39µm(LL6) and 46µm(LR6). The repeatability was significantly poorer for LL6 and LR6, compared to LL1 using Levene’s Test (p=0.003 LL1,LR6 and p=0.007 LL1,LL6). The accuracy of the calculated distances moved by LL1, LL6 and LR6 was -15(22)µm, -93(121)µm and 4(53)µm [mean(standard deviation)].

Conclusions: Optical interocclusal records produce good levels of accuracy and repeatability. The precision diminishes at points located further from the actual scan, but remains clinically acceptable.