Posterior Approximate Clustering-Based Sensitivity Matrix Decomposition for Electrical Impedance Tom...
Posterior Approximate Clustering-Based Sensitivity Matrix Decomposition for Electrical Impedance Tomography
About this item
Full title
Author / Creator
Publisher
Switzerland: MDPI AG
Journal title
Language
English
Formats
Publication information
Publisher
Switzerland: MDPI AG
Subjects
More information
Scope and Contents
Contents
This paper introduces a sensitivity matrix decomposition regularization (SMDR) method for electric impedance tomography (EIT). Using
-means clustering, the EIT-reconstructed image can be divided into four clusters, derived based on image features, representing posterior information. The sensitivity matrix is then decomposed into distinct work ar...
Alternative Titles
Full title
Posterior Approximate Clustering-Based Sensitivity Matrix Decomposition for Electrical Impedance Tomography
Authors, Artists and Contributors
Author / Creator
Identifiers
Primary Identifiers
Record Identifier
TN_cdi_doaj_primary_oai_doaj_org_article_9e87bf2649a348dfa7fcaf6ceac9c3d2
Permalink
https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_doaj_primary_oai_doaj_org_article_9e87bf2649a348dfa7fcaf6ceac9c3d2
Other Identifiers
ISSN
1424-8220
E-ISSN
1424-8220
DOI
10.3390/s24020333