Stabilisation splints produced by artificial intelligence : A pilot study

Sammanfattning

Background: Annually approximately 50 000 stabilisation splints (SS) are produced in Sweden. Variations in fit of the splints affects the adjustment time needed and may also affect the quality of the treatment. Artificial Intelligence (AI) has been introduced to produce SS. The AI technique may offer more standardized splints. Aims: 1. To evaluate the fit and contact pattern of AI-manufactured SS. 2. To test a trigonometric calculation model of the intermaxillary angle (α) and whether that angle affects the contact pattern of the lower jaw teeth against the SS. 3. To assess the inter- and intra-individual reliability of repeated α measurement. Methods: The study was divided into two parts. First, evaluation forms were collected from the Umeå University dentistry program to assess the fit, contact pattern and adjustment time of AI- produced SS. Secondly, a clinical- and laboratory exploratory study; Five participants were scanned for SS, and the SS was evaluated upon delivery. A calculation model was developed and multiple α measurements of digital representations of the upper and lower jaw in a 3D program assessed intra- and interindividual reliability. Results: The fit, retention and design of the SS’s were generally without remark. Initial contact patterns showed wide variation and only a few SS had optimal initial contact patterns. No significant correlation between α and contact pattern was found. The intra- and interindividual measurements showed excellent reliability. Conclusion: There exists a wide variation in initial contact patterns on AI-produced SS. It is possible to calculate α with high reproducibility.