ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술지 Neural Network Image Reconstruction for Magnetic Particle Imaging
Cited 3 time in scopus Download 54 time Share share facebook twitter linkedin kakaostory
저자
채병규
발행일
201712
출처
ETRI Journal, v.39 no.6, pp.841-850
ISSN
1225-6463
출판사
한국전자통신연구원 (ETRI)
DOI
https://dx.doi.org/10.4218/etrij.2017-0094
협약과제
16MS1400, 차세대 의료영상 이미징 시스템 개발, 홍효봉
초록
We investigate neural network image reconstruction for magnetic particle imaging. The network performance strongly depends on the convolution effects of the spectrum input data. The larger convolution effect appearing at a relatively smaller nanoparticle size obstructs the network training. The trained single-layer network reveals the weighting matrix consisting of a basis vector in the form of Chebyshev polynomials of the second kind. The weighting matrix corresponds to an inverse system matrix, where an incoherency of basis vectors due to low convolution effects, as well as a nonlinear activation function, plays a key role in retrieving the matrix elements. Test images are well reconstructed through trained networks having an inverse kernel matrix. We also confirm that a multi-layer network with one hidden layer improves the performance. Based on the results, a neural network architecture overcoming the low incoherence of the inverse kernel through the classification property is expected to become a better tool for image reconstruction.
키워드
Convolution effects, Image reconstruction, Magnetic particle imaging, Neural network, System matrix
KSP 제안 키워드
Chebyshev polynomials of the second kind, Hidden layer, Image reconstruction, Inverse system, Key role, Matrix element, Multi-layer network, Nanoparticle size, Network Image, Network performance, Network training
본 저작물은 공공누리 제4유형 : 출처표시 + 상업적 이용금지 + 변경금지 조건에 따라 이용할 수 있습니다.
제4유형