AI Alignment of Maxillofacial Volumetric and Surface Scans
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing dental imaging technologies, such as cone beam computed tomography (CBCT), are time-consuming and require significant personnel training for accurate interpretation, and the manual alignment of volumetric and surface scans is labor-intensive, hindering efficient dental diagnostics.
Innovation Solution
An automated parsing pipeline system using AI/ML models, including V-Net and DenseNet convolutional neural networks, for parsing, localization, and alignment of volumetric and surface scan images, enabling automated anatomical structure identification and condition classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual alignment of volumetric and surface scans is performed, then alignment accuracy can be achieved, but time consumption and labor intensity increase significantly
Solution Approach 1:
The patent replaces the manual mechanical alignment process with an automated AI-based system. The AI model automatically processes volumetric CBCT images and surface scan images, extracting anatomical structures and performing alignment without human intervention, thereby eliminating time-consuming manual operations while maintaining accuracy
Solution Approach 2:
The system performs self-alignment by automatically processing and aligning the volumetric and surface scan images itself. The AI model extracts features from both image types, identifies corresponding anatomical structures, and computes the optimal alignment transformation, allowing the system to serve its own alignment needs without requiring manual operator intervention
2Loss of information
If CBCT imaging is used, then 3D anatomical information is obtained, but personnel training requirements and operational complexity increase
Solution Approach 1:
The patent replaces complex manual CBCT processing operations with automated AI-based image processing. The AI model automatically segments anatomical structures, extracts 3D information, and performs analysis without requiring operators to manually navigate complex imaging software or interpret raw DICOM data, thereby reducing operational complexity
Solution Approach 2:
The AI model acts as an intermediary between the raw CBCT imaging data and the final diagnostic output. It automatically processes the volumetric data, extracts anatomical information, and presents processed results in a user-friendly format, shielding operators from the complexity of raw image processing while preserving complete 3D anatomical information
3Measurement precision
If deep learning AI models are applied to interpret images, then diagnostic accuracy is improved, but computational resources and processing time increase
Solution Approach 1:
The patent segments the complex diagnostic task into multiple specialized AI models, each handling specific anatomical regions or diagnostic objectives. By dividing the volumetric and surface scan processing into separate modular models, the system can process data in parallel and reduce overall computational load while maintaining high diagnostic accuracy through specialized processing
Data Source
AI summary
A method for alignment of volumetric and surface scan images, said method comprising the steps of: receiving a volumetric image and surface scan image, wherein the volumetric image is a three-dimensional voxel array of a maxillofacial anatomy of a patient and the surface scan image is a polygonal mesh corresponding to the maxillofacial anatomy of the same patient; segmenting the volumetric image and surface scan image into a set of distinct anatomical structures by assigning each voxel in the volumetric image an identifier by structure and assigning each vertex or face of the mesh from the surface scan image an identifier by structure, wherein at least one of the distinct anatomical structures are in common between the volumetric and the surface scan image; extracting a polygonal mesh from the volumetric image featuring common structures with the polygonal mesh from the surface scan image; converting both meshes from the volumetric image and from the surface scan to a point cloud; and aligning the converted meshes via point clouds using a point set registration.