Polar Coordinate Image Segmentation for Medical Lumen Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Medical images expressed in polar coordinates, when transformed to rectangular coordinates, often result in discontinuous portions, leading to inaccurate prediction of boundary positions, particularly in segmentation tasks using convolutional neural networks.

Innovation Solution

An information processing device and method that acquires polar coordinate medical images, classifies image regions using a trained model, and transforms the segment data from polar to rectangular coordinates, ensuring accurate prediction of image regions by extending the input angle beyond 360 degrees to account for discontinuities.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If a polar coordinate image is transformed to a rectangular coordinate image, then the image can be displayed in real space, but discontinuous portions are introduced leading to inaccurate boundary prediction

Engineering Contradiction:

Improveimage display capabilityVSAvoidboundary position prediction accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent applies preliminary action by extending the polar coordinate image to cover more than 360 degrees before transformation. This pre-processing step ensures that the discontinuity problem is prevented in advance by providing overlapping angular information, allowing the segmentation model to accurately predict boundary positions even after transformation to rectangular coordinates.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces an additional dimensional aspect by extending the angular range beyond the traditional 360-degree limit. This dimensional extension creates redundant information across the discontinuity boundary, enabling the model to maintain measurement precision while still allowing transformation to rectangular coordinates for display purposes.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Ease of manufacture

If segmentation is performed after transforming polar coordinate image to rectangular coordinate image, then the process follows conventional workflow, but boundary position cannot be accurately predicted due to discontinuous portions

Engineering Contradiction:

Improveprocessing workflow simplicityVSAvoidboundary position prediction accuracy

Core Design Contradiction:

Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent performs the extension of polar coordinate image coverage to more than 360 degrees as a preliminary action before segmentation. This ensures that when segmentation is subsequently performed on the extended image, the model has access to continuous information across all boundary positions, thereby maintaining both workflow simplicity and prediction accuracy.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent applies segmentation to the extended polar coordinate image that covers more than 360 degrees. By segmenting this extended image rather than a traditional 360-degree image, the method captures continuous boundary information across all angular positions, eliminating the discontinuity problem while maintaining a straightforward segmentation workflow.

Inventive Principle:

Principle #1Segmentation

3Area of stationary object

If the polar coordinate image covers exactly 360 degrees, then the complete circumferential view is provided, but discontinuities occur at the boundary leading to prediction errors

Engineering Contradiction:

Improvecircumferential coverageVSAvoidprediction accuracy at boundary

Core Design Contradiction:

Area of stationary objectVSReliability

Solution Approach 1:

The patent applies preliminary action by extending the polar coordinate image coverage to more than 360 degrees before any processing occurs. This pre-extension creates overlapping angular information that eliminates boundary discontinuities, ensuring reliable predictions at all circumferential positions while maintaining complete coverage of the biological lumen.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent ensures continuity of useful action by extending the angular coverage beyond 360 degrees, creating a continuous information field that eliminates discontinuities at the boundary. This continuous coverage allows the segmentation model to reliably predict boundary positions throughout the entire circumferential view without errors at transition points.

Inventive Principle:

Principle #20Continuity of useful action

Data Source

PatentUS20230237657A1Information processing device, information processing method, program, model generating method, and training data generating method

Publication Date: 2023.07.27 TERUMO KK

AI summary

An information processing device configured to: acquire a polar coordinate image, which is a medical image expressed in polar coordinates and obtained by imaging a biological lumen with a device configured to be inserted into the biological lumen, the polar coordinate image having a first axis representing an angle and a second axis intersecting the first axis and representing a distance from the device; input the polar coordinate image for a predetermined angle exceeding 360 degrees to a model trained, when the polar coordinate image is input, to output first segment data in which an image region corresponding to a specific object and another image region are classified, and output the first segment data for the predetermined angle; extract the first segment data for 360 degrees from the first segment data for the predetermined angle; and transform the extracted first segment data to second segment data expressed in rectangular coordinates.