3D Model Slice-Based ML Reconstruction for Spatial Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 3D data processing methods are resource-intensive and prone to human error, particularly in applications like dental diagnostics and spine measurements, due to their high dimensionality and the challenges of obtaining comprehensive spatial information.
Innovation Solution
Computing a central axis for a 3D model and generating multiple 2D slices that intersect this axis, using a trained ML model to process each slice and convert back to 3D information, optimizing the number of slices for training and processing to ensure comprehensive coverage without redundancy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If 3D data is processed directly using traditional methods, then comprehensive spatial information is obtained, but computational resources are excessively consumed and processing time increases
Solution Approach 1:
The patent segments the 3D data into multiple 2D slices that intersect along a central axis. Each slice is processed independently through the ML model, dividing the computationally intensive 3D processing task into manageable 2D sub-tasks. This segmentation maintains spatial information accuracy while significantly reducing computational resource consumption and processing time.
Solution Approach 2:
The patent transforms the 3D processing problem into a 2D problem by creating slices at a different dimensional level. By processing 2D slices and then reconstructing 3D information through coordinate transformation, the system achieves comparable spatial accuracy with reduced computational complexity, as 2D ML processing is less resource-intensive than 3D processing.
2Measurement precision
If the number of 2D slices is increased to improve coverage of the 3D model, then comprehensive spatial information is achieved, but processing time and computational load increase
Solution Approach 1:
The patent applies partial action by selecting a specific number of 2D slices that provides sufficient coverage of the 3D model without processing every possible slice. The slices are strategically positioned to capture essential spatial features while avoiding redundant processing, achieving comprehensive coverage accuracy with optimized processing time.
Solution Approach 2:
The patent performs preliminary action by pre-determining the optimal number and positioning of 2D slices before processing. The central axis is computed in advance, and slice locations are planned to maximize coverage efficiency, preventing unnecessary processing time while ensuring comprehensive spatial information is captured.
3Measurement precision
If manual processing of 3D data is performed to ensure accuracy, then detailed spatial analysis is achieved, but human error increases and productivity decreases
Solution Approach 1:
The patent replaces manual mechanical processing with an automated ML-based system. The ML model processes 2D slices objectively without human intervention, eliminating human error while maintaining diagnostic accuracy. The automated system consistently applies the same processing criteria across all data, improving reliability compared to manual methods.
Solution Approach 2:
The patent enables self-service by allowing the ML model to automatically perform the entire processing pipeline from 2D slice analysis to 3D reconstruction without human intervention. The system independently extracts features, makes diagnoses, and generates results, eliminating human error while maintaining high diagnostic accuracy through the model's learned patterns.
Data Source
AI summary
Computing a central axis for a three-dimensional (3D) model, generating a number of two-dimensional (2D) slices from the 3D model, the central axis passing through each of the 2D slices, and 2D points of the plurality of 2D slices corresponding to 3D points from the 3D model via a 2D point-3D point correspondence. The method also includes computing 3D information about the 3D model by proposing for each 2D slice of the plurality of 2D slices, using a trained ML model, 2D information about the 2D slice using the 2D slice as input, to obtain a plurality of 2D information for the plurality of 2D slices, and converting the plurality of 2D information to the 3D information about the 3D model based on the 2D point-3D point correspondence.