FD-OCT Noise Reduction via Adjacent Depth Data Subtraction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Frequency domain optical coherence tomography (FD-OCT) images are susceptible to fixed horizontal stripe noise due to light source fluctuations, particularly in dynamic environments like medical endoscopy, where conventional noise reduction methods fail to effectively reduce noise without additional hardware or complex calculations.

Innovation Solution

An information processing apparatus and method that reduces noise by subtracting depth information from adjacent measurement times, using a simple matrix subtraction operation, allowing for efficient noise reduction even with unstable light sources without altering hardware configurations or performing additional hardware calculations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If conventional noise reduction methods (mean-line or median-line subtraction) are used, then fixed pattern noise can be reduced under stable light conditions, but noise reduction fails when light source intensity fluctuates dynamically

Engineering Contradiction:

Improvefixed pattern noiseVSAvoidnoise reduction effectiveness

Core Design Contradiction:

Object-affected harmful factorsVSReliability

Solution Approach 1:

The patent transitions from static noise reduction methods (mean-line/median-line subtraction that assume stable light conditions) to a dynamic method that processes depth information from multiple adjacent A-scans. By dynamically adapting to light source fluctuations through real-time processing of neighboring scan data, the system maintains noise reduction effectiveness under varying illumination conditions.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent performs preliminary processing by collecting depth information from multiple adjacent A-scans before conducting the subtraction operation. This preliminary data gathering allows the system to prepare for potential light source fluctuations by having adjacent scan data ready, enabling effective noise reduction even when intensity changes occur during measurement.

Inventive Principle:

Principle #10Preliminary action

2Object-affected harmful factors

If reference spectrum subtraction method is used, then noise can be reduced, but additional hardware operations and pre-scanning are required

Engineering Contradiction:

ImprovenoiseVSAvoidhardware operations

Core Design Contradiction:

Object-affected harmful factorsVSDevice complexity

Solution Approach 1:

The patent extracts and removes the DC spectrum component that causes fixed pattern noise by subtracting depth information from adjacent A-scans. This extraction approach eliminates the need for separate reference spectrum acquisition hardware and operations, achieving noise reduction through software-based processing of existing depth data.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent replaces the mechanical/optical reference spectrum subtraction method with a computational approach. Instead of using additional hardware to obtain and subtract reference spectra, the system uses software-based subtraction of depth information from adjacent A-scans, simplifying the hardware requirements while maintaining noise reduction capability.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Ease of operation

If mean-line or median-line subtraction is used, then software-based noise reduction is achieved, but performance degrades with optical fiber-based systems experiencing light fluctuations

Engineering Contradiction:

Improvesoftware-based processingVSAvoidnoise from light fluctuations

Core Design Contradiction:

Ease of operationVSObject-affected harmful factors

Solution Approach 1:

The patent enhances software-based processing by implementing a dynamic noise reduction algorithm that processes depth information from multiple adjacent A-scans. This dynamic approach adapts to light source fluctuations in real-time, maintaining effectiveness in optical fiber-based systems where intensity variations occur due to bending, heartbeat, breathing, or blood circulation.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent ensures continuous noise reduction by processing depth information from multiple adjacent A-scans in sequence. This continuous processing approach maintains consistent performance throughout the measurement period, even when light source intensity fluctuates dynamically, ensuring the useful action of noise reduction persists throughout the entire imaging process.

Inventive Principle:

Principle #20Continuity of useful action

Data Source

PatentUS9271629B2Frequency domain optical coherence tomography image noise reduction by subtraction of adjacent depth data

Publication Date: 2016.03.01 KOREA UNIV RES & BUSINESS FOUND

AI summary

An information processing apparatus for frequency domain optical coherence tomography comprises an image scanning unit that receives measurement data from an optical coherence tomography device, while moving a measurement point to the next adjacent point at each measurement time by controlling the optical coherence tomography device, and generates depth information based on the measurement data to produce a raw tomographic image. The apparatus includes a noise reduction unit that reduces noise from the raw tomographic image to produce a result tomographic image. The noise reduction unit reduces the noise from the raw tomographic image by subtracting the depth information generated by the image scanning unit at each measurement time from the depth information generated at another measurement time adjacent to said each measurement time.