Presentation | 2019-09-05 [Short Paper] Dynamic PET Image Reconstruction using Non-Negative Matrix Decomposition with Deep Image Prior Tomoshige Shimomura, Kazuya Kawai, Muneyuki Sakata, Yuichi Kimura, Tatsuya Yokota, Hidekata Hontani, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | We present a PET image reconstruction method that can reconstruct dynamic PET images with high SN ratio and can simultaneously segment the images into regions each of which has a different ligand dynamics. It is known that PET image reconstruction is highly sensitive with measurement noises included in sinograms and that stable PET image reconstruction needs appropriate constraints over the solution space. The proposed method constrains the temporal patterns of each voxel using the quadratic variation and the spatial patterns using a Deep Image Prior. For introducing these constraints of the solution space, the proposed method represents each of the temporal series of sinograms and of corresponding PET images using a matrix and factorizes the latter matrix into bases of spatial and temporal patterns. Here, the non-negative constraint is also introduced in the factorization. This non-negative matrix factorization enables the segmentation of the spatial patterns based on the difference of the temporal patterns. In this presentation, we describe the algorithm and show some experimental results. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | PET image / Image reconstruction / Deep Image Prior / Non-negative matrix factorization |
Paper # | PRMU2019-24,MI2019-43 |
Date of Issue | 2019-08-28 (PRMU, MI) |
Conference Information | |
Committee | PRMU / MI / IPSJ-CVIM |
---|---|
Conference Date | 2019/9/4(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | Yoichi Sato(Univ. of Tokyo) / Yoshiki Kawata(Tokushima Univ.) |
Vice Chair | Toru Tamaki(Hiroshima Univ.) / Akisato Kimura(NTT) / Takayuki Kitasaka(Aichi Inst. of Tech.) / Hidekata Hontani(Nagoya Inst. of Tech.) |
Secretary | Toru Tamaki(NTT) / Akisato Kimura(OMRON SINICX) / Takayuki Kitasaka(Yamaguchi Univ.) / Hidekata Hontani(Univ. of Hyogo) |
Assistant | Yusuke Uchida(DeNA) / Takayoshi Yamashita(Chubu Univ.) / Hotaka Takizawa(Tsukuba Univ.) / Yoshito Otake(NAIST) |
Paper Information | |
Registration To | Technical Committee on Pattern Recognition and Media Understanding / Technical Committee on Medical Imaging / Special Interest Group on Computer Vision and Image Media |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | [Short Paper] Dynamic PET Image Reconstruction using Non-Negative Matrix Decomposition with Deep Image Prior |
Sub Title (in English) | |
Keyword(1) | PET image |
Keyword(2) | Image reconstruction |
Keyword(3) | Deep Image Prior |
Keyword(4) | Non-negative matrix factorization |
1st Author's Name | Tomoshige Shimomura |
1st Author's Affiliation | Nagoya Institute of Technology(NIT) |
2nd Author's Name | Kazuya Kawai |
2nd Author's Affiliation | Nagoya Institute of Technology(NIT) |
3rd Author's Name | Muneyuki Sakata |
3rd Author's Affiliation | Tokyo Metropolitan Institute of Gerontology(Tokyo Metro. Inst. Gerontology) |
4th Author's Name | Yuichi Kimura |
4th Author's Affiliation | Kindai University(KU) |
5th Author's Name | Tatsuya Yokota |
5th Author's Affiliation | Nagoya Institute of Technology(NIT) |
6th Author's Name | Hidekata Hontani |
6th Author's Affiliation | Nagoya Institute of Technology(NIT) |
Date | 2019-09-05 |
Paper # | PRMU2019-24,MI2019-43 |
Volume (vol) | vol.119 |
Number (no) | PRMU-192,MI-193 |
Page | pp.pp.69-70(PRMU), pp.69-70(MI), |
#Pages | 2 |
Date of Issue | 2019-08-28 (PRMU, MI) |