3D Maxillary Sinus Imaging Using Coronal-Guided ROI Extraction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for maxillary sinus image processing in digital dentistry are challenging due to the difficulty in distinguishing the maxillary sinus region from other similar cells in axial plane images, leading to noise inclusion and inaccurate image generation.

Innovation Solution

A method and apparatus using first and second artificial neural network models to distinguish coronal and axial plane regions-of-interest, allowing for accurate extraction of a maxillary sinus region from coronal and axial plane images, respectively, and generating a three-dimensional maxillary sinus image.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If a single neural network model is used to directly distinguish the maxillary sinus region from axial plane images, then the processing speed is improved, but the image accuracy deteriorates due to inclusion of noise from similar cells

Engineering Contradiction:

Improveimage processing speedVSAvoidmaxillary sinus region distinction accuracy

Core Design Contradiction:

SpeedVSMeasurement precision

Solution Approach 1:

The patent divides the image processing task into two sequential stages using two different neural network models. The first model processes coronal plane images to identify region-of-interest, while the second model processes axial plane images to extract the maxillary sinus region. This segmentation of the processing task allows each model to specialize in specific aspects, improving overall accuracy while maintaining efficiency through optimized processing at each stage.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent transitions from processing only axial plane images to a two-dimensional approach by incorporating coronal plane images. By adding the coronal plane dimension and using the first neural network model to identify regions of interest in this additional dimension, the system can filter out noise from similar cells that appear in axial plane images, thereby improving distinction accuracy.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Device complexity

If the entire axial plane image is processed to extract the maxillary sinus region, then the processing simplicity is improved, but the image quality deteriorates due to noise inclusion

Engineering Contradiction:

Improveprocessing method simplicityVSAvoidmaxillary sinus image quality

Core Design Contradiction:

Device complexityVSManufacturing precision

Solution Approach 1:

The patent extracts and removes noise from the image processing by using the first neural network model to identify and select only the relevant region-of-interest from the coronal plane image before processing. This extraction of the maxillary sinus region from the broader context of similar cells allows the second neural network model to focus only on the relevant area in the axial plane image, thereby improving image quality while maintaining reasonable processing simplicity.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent performs preliminary processing by using the first neural network model to distinguish the coronal plane region-of-interest before the second neural network model processes the axial plane image. This preliminary action of identifying the region of interest in advance allows the subsequent processing to focus only on the relevant area, improving the final image quality without significantly increasing overall processing complexity.

Inventive Principle:

Principle #10Preliminary action

3Extent of automation

If digital dentistry image processing is applied to maxillary sinus lifting, then the automation level is improved, but the reliability deteriorates due to difficulty in distinguishing similar cells

Engineering Contradiction:

Improveimage processing automationVSAvoidmaxillary sinus region identification accuracy

Core Design Contradiction:

Extent of automationVSReliability

Solution Approach 1:

The patent segments the automated image processing into two specialized stages: first, using a neural network model to process coronal plane images and identify region-of-interest; second, using another neural network model to process axial plane images and extract the maxillary sinus region. This segmentation of the automation process into specialized stages improves reliability by allowing each model to be optimized for its specific task, reducing the difficulty of distinguishing similar cells.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary step where the first neural network model processes coronal plane images to identify regions of interest, which then serve as guidance for the second neural network model's processing of axial plane images. This intermediary role of the first model's output helps the second model accurately distinguish the maxillary sinus region from similar cells, thereby improving overall reliability of the automated processing.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS12505555B2Device and method for providing sinus image and learning method thereof

Publication Date: 2025.12.23 OSSTEMIMPLANT CO LTD

AI summary

An apparatus for providing a maxillary sinus image according to one embodiment includes a first artificial neural network model configured to distinguish a partial coronal plane region-of-interest from a coronal plane image of a head, a second artificial neural network model configured to, relative to an axial plane image of the head, extract a two-dimensional maxillary sinus image in an axial plane region-of-interest that corresponds to a position of the coronal plane region-of-interest, and an image processor configured to generate a three-dimensional maxillary sinus image using the two-dimensional maxillary sinus image.