Software development for the automatic calculation of dental arch relationships on digital models in cleft lip and palate

Abstract

Background: The assessment of dental arch relationships in the evaluation of surgical outcome in orofacial clefting is essential for quality assurance and optimisation of surgical protocols. Valid, reliable and standardised tools are required for multicentre comparison, evidence-based research and audit. Limitations with existing indices combined with emerging digital technology has led to a need for more contemporary, accurate and thorough measures of surgical outcome. 
Objectives: To design, calibrate and validate an automated scoring tool for the assessment of surgical outcomes on digital models, using the modified Huddart and Bodenham (MHB) scoring system.
Design: Retrospective cohort of plaster and digital models from patients with unilateral cleft lip and palate. 
Method: Design and development of an automated software tool took place prior to calibration of three examiners. Fifty-three digital and plaster study models were measured using conventional methods of MHB assessment. The MHB score was also measured using the automated software tool on digital models. The scoring was repeated twice one month apart and intra-observer reliability was calculated for each observer for conventional and automated MHB scoring using intra-class correlation coefficients. Cronbach’s alpha was used to calculate the inter-examiner agreement for scoring. Bland-Altman plots were used to demonstrate the level of agreement for each model medium. 
Results: Intra-observer reliability was excellent for all examiners (0.988, 0.987 & 0.986). Overall agreement using Cronbach’s Alpha for the three methods of scoring was also excellent (0.986). The automated software tool demonstrated the highest inter-observer reliability (0.991), followed by Plaster (0.989), followed by digital models using conventional MHB scoring (0.979). Bland-Altman plots confirmed no systemic bias and greater consistency of the scores with the automated software tool. 
Conclusion: The study presents a valid, reliable and contemporary automatic scoring software tool for the assessment of dental arch relationships in orofacial clefting. The automated scoring tool has superior consistency and reproducibility to conventional methods of assessment using the MHB index.

Date of Award	2015
Original language	English
Awarding Institution	University of Dundee
Supervisor	Grant McIntyre (Supervisor) & Peter Mossey (Supervisor)