High-Fidelity 3D Tooth Reconstruction by Fusing Intraoral Scans and CBCT Data via a Deep Implicit Representation

2026-01-21

AI Summary

Current dental imaging modalities struggle to simultaneously achieve high-fidelity reconstruction of both the crown and root of teeth: intraoral scans provide high accuracy for crowns but lack root information, while cone-beam computed tomography (CBCT) captures roots but yields poor crown quality; direct fusion of these modalities often introduces artifacts and geometric discontinuities. This work proposes a fully automatic, multi-modal fusion framework based on deep implicit representations. By integrating tooth instance segmentation, robust registration, and hybrid proxy mesh construction, the method leverages class-specific DeepSDF networks to project input data onto a manifold of idealized tooth shapes, yielding anatomically consistent, watertight, and seamless 3D dental models. To our knowledge, this is the first approach to apply deep implicit representations to multi-modal dental data fusion, successfully preserving high crown fidelity while fully reconstructing patient-specific roots, significantly outperforming conventional stitching strategies.

Technology Category

• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Multimodal Learning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

High-fidelity 3D tooth models are essential for digital dentistry, but must capture both the detailed crown and the complete root. Clinical imaging modalities are limited: Cone-Beam Computed Tomography (CBCT) captures the root but has a noisy, low-resolution crown, while Intraoral Scanners (IOS) provide a high-fidelity crown but no root information. A naive fusion of these sources results in unnatural seams and artifacts. We propose a novel, fully-automated pipeline that fuses CBCT and IOS data using a deep implicit representation. Our method first segments and robustly registers the tooth instances, then creates a hybrid proxy mesh combining the IOS crown and the CBCT root. The core of our approach is to use this noisy proxy to guide a class-specific DeepSDF network. This optimization process projects the input onto a learned manifold of ideal tooth shapes, generating a seamless, watertight, and anatomically coherent model. Qualitative and quantitative evaluations show our method uniquely preserves both the high-fidelity crown from IOS and the patient-specific root morphology from CBCT, overcoming the limitations of each modality and naive stitching.

Problem

• 3D tooth reconstruction
• CBCT
• Intraoral Scanning
• high-fidelity modeling
• data fusion

Innovation

• deep implicit representation
• tooth reconstruction
• multimodal fusion
• DeepSDF
• digital dentistry