Discrete-MultiResUNet: Segmentation and feature extraction model for knee MR images

Article type: Research Article

Authors: More, Sujeet | Singla, Jimmy

Affiliations: School of Computer Science and Engineering, Lovely Professional University, Jalandhar, India

Correspondence: [*] Corresponding author. Sujeet More, Research Scholar, Computer Science and Engineering, Lovely Professional University Faculty of Technology and Sciences, India. Tel.: +919886301346; E-mail: sujeet.11816272@lpu.in.

Abstract: Deep learning has shown outstanding efficiency in medical image segmentation. Segmentation of knee tissues is an important task for early diagnosis of rheumatoid arthritis (RA) with selecting variant features. Automated segmentation and feature extraction of knee tissues are desirable for faster and reliable analysis of large datasets and further diagnosis. In this paper a novel architecture called as Discrete-MultiResUNet, which is a combination of discrete wavelet transform (DWT) with MultiResUNet architecture is applied for feature extraction and segmentation, respectively. This hybrid architecture captures more prominent features from the knee magnetic resonance image efficiently with segmenting vital knee tissues. The hybrid model is evaluated on the knee MR dataset demonstrating outperforming performance compared with baseline models. The model achieves excellent segmentation performance accuracy of 96.77% with a dice coefficient of 98%.

Keywords: MultiResUNet, discrete wavelet transform, dice similarity coefficient, rheumatoid arthritis, segmentation

DOI: 10.3233/JIFS-211459

Journal: Journal of Intelligent & Fuzzy Systems, vol. 41, no. 2, pp. 3771-3781, 2021