Artificial intelligence-designed single molar dental prostheses: A protocol of prospective experimental study

Abstract

Background: Dental prostheses, which aim to replace missing teeth and to restore patients' appearance and oral functions, should be biomimetic and thus adopt the occlusal morphology and three-dimensional (3D) position of healthy natural teeth. Since the teeth of an individual subject are controlled by the same set of genes (genotype) and are exposed to mostly identical oral environment (phenotype), the occlusal morphology and 3D position of teeth of an individual patient are inter-related. It is hypothesized that artificial intelligence (AI) can automate the design of single-tooth dental prostheses after learning the features of the remaining dentition.

Materials and methods: This article describes the protocol of a prospective experimental study, which aims to train and to validate the AI system for design of single molar dental prostheses. Maxillary and mandibular dentate teeth models will be collected and digitized from at least 250 volunteers. The (original) digitized maxillary teeth models will be duplicated and processed by removal of right maxillary first molars (FDI tooth 16). Teeth models will be randomly divided into training and validation sets. At least 200 training sets of the original and the processed digitalized teeth models will be input into 3D Generative Adversarial Network (GAN) for training. Among the validation sets, tooth 16 will be generated by AI on 50 processed models and the morphology and 3D position of AI-generated tooth will be compared to that of the natural tooth in the original maxillary teeth model. The use of different GAN algorithms and the need of antagonist mandibular teeth model will be investigated. Results will be reported following the CONSORT-AI.

Fig 1. SPIRIT diagram showing the schedule of enrollment, allocation, interventions, and assessments of this study.

–t₁  =  baseline assessment (before randomization), 0 (t₀) = randomization, t₁ = after randomization, t₂  =  25% of collected models learnt by AI, t₃  =  50% of collected models learnt by AI, t₄ = 75% of collected models learnt by AI, t₅  =  100% of collected models learnt by AI, t₆ = the whole study is completed.

Fig 2. General flow of processing collected data.

Fig 3. Brief architecture of the Generative Adversarial Network (GAN) of this study.

Original model: digitized maxillary teeth model collected from participants; processed model: teeth model with tooth 16 removed; Generated model: teeth model with AI-generated tooth 16.

Fig 4. System specification of high-performance computer.

Republished from [31] under a CC BY license, with permission from ITS, HKU, original copyright 2022.

Fig 5

A. Illustration of voxel-based method. B. Illustration of view-based method. C. Illustration of point-based method.

Fig 6

A & B: Example of a maxillary teeth model (left) and a mandibular teeth model (right). C: Demonstration of maxillary teeth model and its antagonist mandibular teeth model at maximal intercuspal position.

Fig 7

A & B: Superimposition of teeth for comparison (left). Measurement of the geometric morphology and 3D position by locating the anatomical landmarks of a tooth such as cusp tips and fossae as well as the center of a tooth (arrowed) respectively (right).