Deep Upscale U-Net for automatic tongue segmentation
Issued Date
2024-01-01
Resource Type
ISSN
01400118
eISSN
17410444
Scopus ID
2-s2.0-85185324399
Journal Title
Medical and Biological Engineering and Computing
Rights Holder(s)
SCOPUS
Bibliographic Citation
Medical and Biological Engineering and Computing (2024)
Suggested Citation
Kusakunniran W., Imaromkul T., Mongkolluksamee S., Thongkanchorn K., Ritthipravat P., Tuakta P., Benjapornlert P. Deep Upscale U-Net for automatic tongue segmentation. Medical and Biological Engineering and Computing (2024). doi:10.1007/s11517-024-03051-w Retrieved from: https://repository.li.mahidol.ac.th/handle/20.500.14594/97358
Title
Deep Upscale U-Net for automatic tongue segmentation
Corresponding Author(s)
Other Contributor(s)
Abstract
Abstract: In a treatment or diagnosis related to oral health conditions such as oral cancer and oropharyngeal cancer, an investigation of tongue’s movements is a major part. In an automatic measurement of such movement, it must first start with a task of tongue segmentation. This paper proposes a solution of tongue segmentation based on a decoder-encoder CNN-based structure i.e., U-Net. However, it could suffer from a problem of feature loss in deep layers. This paper proposes a Deep Upscale U-Net (DU-UNET). An additional up-sampling of the feature map from a contracting path is concatenated to an upper layer of an expansive path, based on an original U-Net structure. The segmentation model is constructed by training DU-UNET on the two publicly available datasets, and transferred to the self-collected dataset of tongue images with five tongue postures which were recorded at a far distance from a camera under a real-world scenario. The proposed DU-UNET outperforms the other existing methods in our literature reviews, with accuracy of 99.2%, mean IoU of 97.8%, Dice score of 96.8%, and Jaccard score of 96.8%. Graphical abstract: (Figure presented.)