K-space Weighted Image Average for Noise Reduction in Medical Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging techniques such as CTP, PET, SPECT, and MRI face challenges with image noise and the need for higher doses of radiation or contrast agents to achieve satisfactory results.
Innovation Solution
The K-space Weighted Image Average (KWIA) method reduces noise and dose by applying view-shared averaging methods to k-space data, preserving spatial and temporal resolution without requiring modifications to existing scanners or lengthy computation times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard image processing techniques are used, then image quality is maintained, but image noise increases and radiation dose or contrast agent dosage must be increased
Solution Approach 1:
The patent introduces k-space as an intermediary domain between raw imaging data and final images. By performing weighted averaging in k-space rather than directly in image space, the method achieves noise reduction while preserving image quality. The k-space representation acts as a mediator that allows selective processing of different spatial frequencies through the weighting function.
Solution Approach 2:
The patent transforms the imaging data from image space to k-space (frequency domain) using Fourier transform, changing the domain parameter. This parameter change enables the application of weighted averaging that would be difficult to implement directly in image space, allowing for effective noise reduction while maintaining image quality.
2Object-affected harmful factors
If radiation dose or contrast agent dosage is reduced, then harmful effects are decreased, but image noise increases and image quality deteriorates
Solution Approach 1:
The k-space domain serves as an intermediary that enables noise reduction processing. By working in this intermediate domain with the weighted averaging function, the patent can compensate for the reduced signal quality resulting from lower radiation or contrast agent doses, thereby maintaining image quality despite reduced dosing.
Solution Approach 2:
The weighting function applied in k-space provides local quality enhancement by differentially processing different regions of k-space. The weighting function w(k) allows certain spatial frequency components to be emphasized or suppressed, effectively enhancing image quality in specific regions while managing the overall noise level resulting from reduced dosing.
3Object-affected harmful factors
If noise reduction techniques are applied, then image noise decreases, but computation time increases and scanner modifications are required
Solution Approach 1:
The patent makes the noise reduction method universally applicable to multiple imaging modalities (CT, PET, SPECT, MRI) without requiring modality-specific implementations or scanner modifications. The k-space weighted averaging approach works across different imaging types, eliminating the need for separate development and validation for each modality, thus reducing overall computation and implementation time.
Solution Approach 2:
By using k-space as an intermediary domain, the patent creates a unified processing framework that works across different imaging modalities. This intermediary representation allows the same noise reduction technique to be applied universally without requiring modality-specific scanner modifications or complex computation tailored to each imaging type.
Data Source
AI summary
Reducing noise and dose (radiation or contrast) for perfusion imaging in Computed Tomography Perfusion (CTP), Positron Emission Tomography (PET), Single Photon Emission Computed Tomography (SPECT), and Magnetic Resonance Imaging (MRI) medical scanning devices by using a k-space based method. The time sequence of images from the scanner data set is converted as necessary, such as using a 2D Fast Fourier Transform (FFT), into a k-space having multiple timeframes. View-shared averaging is performed to reduce noise and preserve spatial and temporal resolutions of CTP, PET, SPECT and MRI data by progressively increasing the number of time frames for view-shared averaging for more distant regions of “k-space”, before converting the data, such as through a 2D FFT into a time sequence of noise reduced images.