D-UNet: A Dimension-Fusion U Shape Network for Chronic Stroke Lesion Segmentation

• Published in: IEEE/ACM transactions on computational biology and bioinformatics Vol. 18; no. 3; pp. 940 - 950
• Main Authors: Zhou, Yongjin, Huang, Weijian, Dong, Pei, Xia, Yong, Wang, Shanshan
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.05.2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Artificial neural networks; Chronic Disease; Coders; Computer applications; Convolution; deep learning; dimensional fusion; Encoders-Decoders; Feature extraction; Humans; Image Interpretation, Computer-Assisted - methods; Image processing; Image segmentation; Imaging, Three-Dimensional; Lesions; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Medical imaging; MRI; Neural networks; Neural Networks, Computer; Propagation losses; Stroke - diagnostic imaging; stroke segmentation; Three-dimensional displays; Two dimensional displays
• ISSN: 1545-5963; 1557-9964; 1557-9964
• DOI: 10.1109/TCBB.2019.2939522

Assessing the location and extent of lesions caused by chronic stroke is critical for medical diagnosis, surgical planning, and prognosis. In recent years, with the rapid development of 2D and 3D convolutional neural networks (CNN), the encoder-decoder structure has shown great potential in the field of medical image segmentation. However, the 2D CNN ignores the 3D information of medical images, while the 3D CNN suffers from high computational resource demands. This paper proposes a new architecture called dimension-fusion-UNet (D-UNet), which combines 2D and 3D convolution innovatively in the encoding stage. The proposed architecture achieves a better segmentation performance than 2D networks, while requiring significantly less computation time in comparison to 3D networks. Furthermore, to alleviate the data imbalance issue between positive and negative samples for the network training, we propose a new loss function called Enhance Mixing Loss (EML). This function adds a weighted focal coefficient and combines two traditional loss functions. The proposed method has been tested on the ATLAS dataset and compared to three state-of-the-art methods. The results demonstrate that the proposed method achieves the best quality performance in terms of DSC = 0.5349 <inline-formula><tex-math notation="LaTeX">\pm</tex-math> <mml:math><mml:mo>±</mml:mo></mml:math><inline-graphic xlink:href="huang-ieq1-2939522.gif"/> </inline-formula> 0.2763 and precision = 0.6331 <inline-formula><tex-math notation="LaTeX">\pm</tex-math> <mml:math><mml:mo>±</mml:mo></mml:math><inline-graphic xlink:href="huang-ieq2-2939522.gif"/> </inline-formula> 0.295).