TY - GEN
T1 - Classification and Recognition of Knee Coronal Femoral Edema Based on Nonnegative Matrix Factorization
AU - Li, Guanyi
AU - Lv, Songcen
AU - Qin, Yong
AU - Jiang, Yuchen
AU - Luo, Hao
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.
PY - 2024
Y1 - 2024
N2 - It’s a huge workload to label the edema area in the diagnosis of knee edema at present, based on the excellent ability of nonnegative matrix factorization to process high-dimensional data and the characteristics of less requirement for prior information, we combine nonnegative matrix factorization with the diagnosis of knee edema to determine whether there is edema and find the corresponding edema area. In this paper, we propose a new weakly-supervised method named FSNMF algorithm, by adjusting the corresponding eigenvalues based on different targets, so as to ensure the effectiveness of disease image classification and edema region feature extraction. In our paper, experiments have been carried out on the proposed algorithm to show the performance of our FSNMF algorithm. Through experiments on public data sets, we verify the classification accuracy of the proposed method is improved compared with the previous NMF classification algorithm and the accuracy of edema identification using FSNMF is also excellent.
AB - It’s a huge workload to label the edema area in the diagnosis of knee edema at present, based on the excellent ability of nonnegative matrix factorization to process high-dimensional data and the characteristics of less requirement for prior information, we combine nonnegative matrix factorization with the diagnosis of knee edema to determine whether there is edema and find the corresponding edema area. In this paper, we propose a new weakly-supervised method named FSNMF algorithm, by adjusting the corresponding eigenvalues based on different targets, so as to ensure the effectiveness of disease image classification and edema region feature extraction. In our paper, experiments have been carried out on the proposed algorithm to show the performance of our FSNMF algorithm. Through experiments on public data sets, we verify the classification accuracy of the proposed method is improved compared with the previous NMF classification algorithm and the accuracy of edema identification using FSNMF is also excellent.
KW - Edema identification
KW - Image classification
KW - Nonnegative matrix factorization
KW - Weakly-supervised method
UR - https://www.scopus.com/pages/publications/85203135159
U2 - 10.1007/978-3-031-67192-0_27
DO - 10.1007/978-3-031-67192-0_27
M3 - 会议稿件
AN - SCOPUS:85203135159
SN - 9783031671913
T3 - Lecture Notes in Networks and Systems
SP - 216
EP - 224
BT - Intelligent and Fuzzy Systems - Intelligent Industrial Informatics and Efficient Networks Proceedings of the INFUS 2024 Conference
A2 - Kahraman, Cengiz
A2 - Cevik Onar, Sezi
A2 - Cebi, Selcuk
A2 - Oztaysi, Basar
A2 - Ucal Sari, Irem
A2 - Tolga, A. Cagrı
PB - Springer Science and Business Media Deutschland GmbH
T2 - International Conference on Intelligent and Fuzzy Systems, INFUS 2024
Y2 - 16 July 2024 through 18 July 2024
ER -