ABSTRACT
In order to increase the segmentation impact of the Diabetic Foot Ulcer Challenge 2022 dataset, we train a selection of popular deep learning segmentation algorithms and improve training methods, such as adding Dice term to loss function, employing transfer learning and poly learning rate update strategy, etc., in this paper. Experiments show that our method is effective, we get a Dice score of 0.7045, which is better than the official baseline result of 0.6277. Moreover, we integrate the above segmentation models using four ensemble methods to evaluate segmentation performance, such as Averaging, Weighting, Voting, and Stacking. We observed that our proposed one-layer CNN stacking network exhibits superior segmentation performance (Dice score: 0.7142) compared to single CNN model and other three ensemble methods. Our performance surpasses the baseline result, placing us in the top 10 in the Diabetic Foot Ulcer Challenge 2022.
- H. Sun, P. Saeedi, S. Karuranga, 2022. IDF Diabetes Atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes research and clinical practice 183, 109119.Google Scholar
- D. G. Armstrong, A. J. M. Boulton, and S. A. Bus. 2017. Diabetic foot ulcers and their recurrence. New England Journal of Medicine 376, 24, 2367–2375.Google ScholarCross Ref
- K. Ogurtsova, S. Morbach, B. Haastert, 2021. Cumulative long-term recurrence of diabetic foot ulcers in two cohorts from centres in Germany and the Czech Republic. Diabetes research and clinical practice 172, 108621.Google Scholar
- E. Ghanassia, L. Villon, 2008. Long-term outcome and disability of diabetic patients hospitalized for diabetic foot ulcers: a 6.5-year follow-up study. Diabetes care 31, 7, 1288–1292.Google ScholarCross Ref
- B. Cassidy, N. D. Reeves, J. M. Pappachan, 2022. A cloud-based deep learning framework for remote detection of diabetic foot ulcers. IEEE Pervasive Computing 21, 2, 78–86.Google ScholarDigital Library
- R. Brown, B. Ploderer, L. S. D. Seng, 2017. MyFootCare: a mobile self-tracking tool to promote self-care amongst people with diabetic foot ulcers. Australian Conference on Computer-Human Interaction, 462–466.Google ScholarDigital Library
- C. Kendrick, B. Cassidy, J. M. Pappachan, 2022. Translating clinical delineation of diabetic foot ulcers into machine interpretable segmentation. arXiv preprint arXiv:2204.11618.Google Scholar
- O. Ronneberger, P. Fischer and T. Brox. 2015. U-net: Convolutional networks for biomedical image segmentation. International Conference on Medical image computing and computer-assisted intervention, 234–241.Google Scholar
- L.-C. Chen, G. Papandreou, F. Schroff, 2017. Rethinking atrous convolution for semantic image segmentation. arXiv preprint arXiv:1706.05587.Google Scholar
- L.-C. Chen, G. Papandreou, I. Kokkinos, 2017. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs. IEEE transactions on pattern analysis and machine intelligence 40, 4, 834–848.Google Scholar
- L.-C. Chen, Y. Zhu, G. Papandreou, 2018. Encoder-decoder with atrous separable convolution for semantic image segmentation. European conference on computer vision, 801–818.Google ScholarDigital Library
- O. Sagi and L. Rokach. 2018. Ensemble learning: A survey. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery 8, 4, e1249.Google ScholarCross Ref
- Y. Zheng, C. Li, X. Zhou, 2022. Application of transfer learning and ensemble learning in image-level classification for breast histopathology. arXiv preprint arXiv:2204.08311.Google Scholar
- S. Rajaraman, F. Yang, G. Zamzmi, 2022. Deep ensemble learning for segmenting tuberculosis-consistent manifestations in chest radiographs. arXiv preprint arXiv:2206.06065.Google Scholar
- M. Goyal, M. H. Yap, N. D. Reeves, 2017. Fully convolutional networks for diabetic foot ulcer segmentation. IEEE international conference on systems, man, and cybernetics, 618–623.Google ScholarDigital Library
- C. Wang, A. Mahbod, I. Ellinger, 2022. FUSeg: The foot ulcer segmentation challenge. arXiv preprint arXiv:2201.00414.Google Scholar
- A. Mahbod, R. Ecker, and I. Ellinger. 2021. Automatic foot ulcer segmentation using an ensemble of convolutional neural networks. arXiv preprint arXiv:2109.01408.Google Scholar
- R. Ribani and M. Marengoni. 2019. A survey of transfer learning for convolutional neural networks. SIBGRAPI Conference on Graphics, Patterns and Images Tutorials, 47–57.Google Scholar
- K. Li, J. Yin, Z. Lu, 2012. Multiclass boosting SVM using different texture features in HEp-2 cell staining pattern classification. International Conference on Pattern Recognition, 170–173.Google Scholar
- A. Howard, M. Sandler, G. Chu, 2019. Searching for mobilenetv3. IEEE/CVF international conference on computer vision, 1314–1324.Google ScholarCross Ref
- K. Simonyan and A. Zisserman. 2014. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556.Google Scholar
- K. He, X. Zhang, S. Ren, 2016. Deep residual learning for image recognition. IEEE conference on computer vision and pattern recognition, 770–778.Google ScholarCross Ref
- J. Wang, K. Sun, T. Cheng, 2020. Deep high-resolution representation learning for visual recognition. IEEE transactions on pattern analysis and machine intelligence 43, 10, 3349–3364.Google Scholar
Index Terms
- Ensemble Learning for Diabetic Foot Ulcer Segmentation based on DFUC2022 Dataset
Recommendations
Diabetic Foot Ulcer Segmentation Using Convolutional and Transformer-Based Models
Diabetic Foot Ulcers Grand ChallengeAbstractDiabetes is a rising global epidemic, it was estimated that in 2017 there are 451 million (aged 18–99 years) people with diabetes worldwide, and it is expected to increase to 693 million by 2045. Diabetic Foot Ulcers (DFU) is a serious disease ...
A Superpixel-Wise Fully Convolutional Neural Network Approach for Diabetic Foot Ulcer Tissue Classification
Pattern Recognition. ICPR International Workshops and ChallengesAbstractAccurate assessment of diabetic foot ulcers (DFU) is primordial to provide an efficient treatment and to prevent amputation. Traditional DFU assessment methods used by clinicians are based on visual examination of the ulcer by estimating the ...
Application of Deep Learning Autoencoders as Features Extractor of Diabetic Foot Ulcer Images
Intelligent Computing MethodologiesAbstractDiabetic Foot Ulcer is one of the most common diabetic complications that can lead to amputation if not treated appropriately and timely. When diagnosed by professionals, diabetic foot ulcers can be extremely successful, but the diagnosis comes at ...
Comments