Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormer
The semantic segmentation of bone structures demands pixel-level classification accuracy to create reliable bone models for diagnosis. While Convolutional Neural Networks (CNNs) are commonly used for segmentation, they often struggle with complex shapes due to their focus on texture features and lim...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/25/1/81 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850116875693850624 |
|---|---|
| author | Naohiro Masuda Keiko Ono Daisuke Tawara Yusuke Matsuura Kentaro Sakabe |
| author_facet | Naohiro Masuda Keiko Ono Daisuke Tawara Yusuke Matsuura Kentaro Sakabe |
| author_sort | Naohiro Masuda |
| collection | DOAJ |
| description | The semantic segmentation of bone structures demands pixel-level classification accuracy to create reliable bone models for diagnosis. While Convolutional Neural Networks (CNNs) are commonly used for segmentation, they often struggle with complex shapes due to their focus on texture features and limited ability to incorporate positional information. As orthopedic surgery increasingly requires precise automatic diagnosis, we explored SegFormer, an enhanced Vision Transformer model that better handles spatial awareness in segmentation tasks. However, SegFormer’s effectiveness is typically limited by its need for extensive training data, which is particularly challenging in medical imaging, where obtaining labeled ground truths (GTs) is a costly and resource-intensive process. In this paper, we propose two models and their combination to enable accurate feature extraction from smaller datasets by improving SegFormer. Specifically, these include the data-efficient model, which deepens the hierarchical encoder by adding convolution layers to transformer blocks and increases feature map resolution within transformer blocks, and the FPN-based model, which enhances the decoder through a Feature Pyramid Network (FPN) and attention mechanisms. Testing our model on spine images from the Cancer Imaging Archive and our own hand and wrist dataset, ablation studies confirmed that our modifications outperform the original SegFormer, U-Net, and Mask2Former. These enhancements enable better image feature extraction and more precise object contour detection, which is particularly beneficial for medical imaging applications with limited training data. |
| format | Article |
| id | doaj-art-ca9366701cc74229a2fde5c4dab3c2af |
| institution | OA Journals |
| issn | 1424-8220 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj-art-ca9366701cc74229a2fde5c4dab3c2af2025-08-20T02:36:12ZengMDPI AGSensors1424-82202024-12-012518110.3390/s25010081Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormerNaohiro Masuda0Keiko Ono1Daisuke Tawara2Yusuke Matsuura3Kentaro Sakabe4Master’s Program in Information and Computer Science, Doshisha University, Kyoto 610-0394, JapanDepartment of Intelligent Information Engineering and Sciences, Doshisha University, Kyoto 610-0394, JapanDepartment of Advanced Science and Technology, Ryukoku University, Kyoto 520-2194, JapanDepartment of Orthopedic Surgery, Chiba University, Chiba 260-8677, JapanMaster’s Program in Information and Computer Science, Doshisha University, Kyoto 610-0394, JapanThe semantic segmentation of bone structures demands pixel-level classification accuracy to create reliable bone models for diagnosis. While Convolutional Neural Networks (CNNs) are commonly used for segmentation, they often struggle with complex shapes due to their focus on texture features and limited ability to incorporate positional information. As orthopedic surgery increasingly requires precise automatic diagnosis, we explored SegFormer, an enhanced Vision Transformer model that better handles spatial awareness in segmentation tasks. However, SegFormer’s effectiveness is typically limited by its need for extensive training data, which is particularly challenging in medical imaging, where obtaining labeled ground truths (GTs) is a costly and resource-intensive process. In this paper, we propose two models and their combination to enable accurate feature extraction from smaller datasets by improving SegFormer. Specifically, these include the data-efficient model, which deepens the hierarchical encoder by adding convolution layers to transformer blocks and increases feature map resolution within transformer blocks, and the FPN-based model, which enhances the decoder through a Feature Pyramid Network (FPN) and attention mechanisms. Testing our model on spine images from the Cancer Imaging Archive and our own hand and wrist dataset, ablation studies confirmed that our modifications outperform the original SegFormer, U-Net, and Mask2Former. These enhancements enable better image feature extraction and more precise object contour detection, which is particularly beneficial for medical imaging applications with limited training data.https://www.mdpi.com/1424-8220/25/1/81feature pyramid networkMask2FormerSegFormersemantic segmentationtransformer block |
| spellingShingle | Naohiro Masuda Keiko Ono Daisuke Tawara Yusuke Matsuura Kentaro Sakabe Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormer Sensors feature pyramid network Mask2Former SegFormer semantic segmentation transformer block |
| title | Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormer |
| title_full | Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormer |
| title_fullStr | Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormer |
| title_full_unstemmed | Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormer |
| title_short | Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormer |
| title_sort | data efficient bone segmentation using feature pyramid based segformer |
| topic | feature pyramid network Mask2Former SegFormer semantic segmentation transformer block |
| url | https://www.mdpi.com/1424-8220/25/1/81 |
| work_keys_str_mv | AT naohiromasuda dataefficientbonesegmentationusingfeaturepyramidbasedsegformer AT keikoono dataefficientbonesegmentationusingfeaturepyramidbasedsegformer AT daisuketawara dataefficientbonesegmentationusingfeaturepyramidbasedsegformer AT yusukematsuura dataefficientbonesegmentationusingfeaturepyramidbasedsegformer AT kentarosakabe dataefficientbonesegmentationusingfeaturepyramidbasedsegformer |