Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Fast shift invariant CPD (canonical polyadic decomposition) method applicable to multi-testee fMRI (functional magnetic resonance imaging) data analysis

A multi-subject, shift-invariant technology, applied in the field of medical signal processing, can solve the problems of large data volume, slow running speed, and high memory requirements, achieve good development prospects and improve the effect of separation performance

Active Publication Date: 2019-03-22
CHANGSHA UNIVERSITY OF SCIENCE AND TECHNOLOGY
View PDF13 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In recent years, research on large-scale multi-subject fMRI data (the number of subjects ranges from tens to tens of thousands) has attracted more and more attention, and the spatial dimension of fMRI data is usually very high, such as more than 50,000 brains. Therefore, when the shift-invariant CPD algorithm is applied to multi-subject fMRI data, there are problems of slow running speed and high memory requirements, especially in the process of updating and estimating the time delay of subjects, the amount of data that needs to be stored bigger

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Fast shift invariant CPD (canonical polyadic decomposition) method applicable to multi-testee fMRI (functional magnetic resonance imaging) data analysis
  • Fast shift invariant CPD (canonical polyadic decomposition) method applicable to multi-testee fMRI (functional magnetic resonance imaging) data analysis
  • Fast shift invariant CPD (canonical polyadic decomposition) method applicable to multi-testee fMRI (functional magnetic resonance imaging) data analysis

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0029] A specific embodiment of the present invention will be described in detail below in conjunction with the technical scheme and accompanying drawings.

[0030] There are currently 10 subjects performing the fMRI data collected under the finger tapping task, that is, K=10. Each subject underwent J=165 scans, and each scan collected 53×63×46 whole-brain data, removed data voxels outside the brain, and kept data voxels I=59610 in the brain. Assuming that the number of components sharing SM and TC components is D = 30, the steps of multi-subject fMRI data analysis using the present invention are shown in the accompanying drawings.

[0031]Step 1: Input multi-subject fMRI data

[0032] The second step: initialization. Randomly initialize shared TC components Shared SM ingredients Subject intensity initial subject delay is a zero matrix, the number of iterations iter=0, the relative error Δε iter = 1, calculate the iteration error ε according to formula (1) iter ....

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a fast shift invariant CPD (canonical polyadic decomposition) method applicable to multi-testee fMRI (functional magnetic resonance imaging) data analysis, and belongs to the field of medical signal processing. Based on a shift invariant CPD algorithm, testee shared SM (spatial map) components, shared TC (time course) components and all testee strength are updated and estimated by an alternating least square method, high-dimensional shared SM components and original multi-testee fMRI data are subjected to matrix multiplication and conversion into low-dimensional data without affecting time delay estimation performances, so that algorithm operation is accelerated, and operation memory is reduced. Needed memory is obviously reduced, and task related components of taskportion multi-testee fMRI data can be rapidly and effectively estimated.

Description

technical field [0001] The invention relates to the field of medical signal processing, in particular to an analysis method for multi-subject functional magnetic resonance imaging fMRI (functional magnetic resonance imaging) data. Background technique [0002] Using a magnetic resonance scanner to scan the brains of multiple subjects, the brain function data obtained are called multi-subject fMRI data. Due to its non-invasive and high spatial resolution advantages, fMRI has become an important technology in current brain science research. Multi-subject fMRI data is generally regarded as a three-dimensional tensor, including space dimension, time dimension and subject dimension, which can be processed by tensor decomposition algorithm. CPD (canonical polyadic decomposition) is a typical tensor decomposition algorithm. CPD has a clear physical meaning in the analysis of multi-subject fMRI data. The multi-subject fMRI data is decomposed into shared brain space activation comp...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): A61B5/055A61B5/00
CPCA61B5/0033A61B5/055A61B5/72A61B2576/00
Inventor 邝利丹林秋华龚晓峰丛丰裕
Owner CHANGSHA UNIVERSITY OF SCIENCE AND TECHNOLOGY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products