Accurate, automated classification of radiographic knee osteoarthritis severity using a novel method of deep learning: Plug‑in modules
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Background: Fine-grained classification deals with data with a large degree of similarity such as cat or bird species and similarly, knee osteoarthritis severity classification (Kellgren-Lawrence [KL] grading) is one of the fine-grained classification tasks. Recently, plug-in module (PIM) that can be integrated into convolutional neural network-based or Transformer-based networks has been shown to provide strong discriminative regions for fine-grained classification with results that outperformed the previous deep learning models. PIM utilizes each pixel of an image as independent feature and can subsequently, better classify images with minor differences. It was hypothesized that as one of the fine-grained classification tasks, knee osteoarthritis severity may be well classified using PIMs. The aim of the study was to develop this automated knee osteoarthritis classification model. Methods: A deep learning model that classifies knee osteoarthritis severity of a radiograph was developed utilizing PIMs. A retrospective analysis on prospectively collected data was performed. The model was trained and developed using the Osteoarthritis Initiative dataset and was subsequently tested on an independent dataset, Multicenter Osteoarthritis Study (test set size: 17040). The final deep learning model was designed through an ensemble of four different PIMs. Results: The accuracies of the model were 85%, 46%, 69%, 82%, and 93% for KL grade 0, 1, 2, 3, and 4, respectively with the overall accuracy of 76.0%. Conclusions: The ensemble of PIMs could classify knee osteoarthritis severity using simple radiographs with a fine accuracy. Although improvements will be needed in the future, the model has been proven to have the potential to be clinically useful.
Abstract
Background: Fine-grained classification deals with data with a large degree of similarity such as cat or bird species and similarly, knee osteoarthritis severity classification (Kellgren-Lawrence [KL] grading) is one of the fine-grained classification tasks. Recently, plug-in module (PIM) that can be integrated into convolutional neural network-based or Transformer-based networks has been shown to provide strong discriminative regions for fine-grained classification with results that outperformed the previous deep learning models. PIM utilizes each pixel of an image as independent feature and can subsequently, better classify images with minor differences. It was hypothesized that as one of the fine-grained classification tasks, knee osteoarthritis severity may be well classified using PIMs. The aim of the study was to develop this automated knee osteoarthritis classification model. Methods: A deep learning model that classifies knee osteoarthritis severity of a radiograph was developed utilizing PIMs. A retrospective analysis on prospectively collected data was performed. The model was trained and developed using the Osteoarthritis Initiative dataset and was subsequently tested on an independent dataset, Multicenter Osteoarthritis Study (test set size: 17040). The final deep learning model was designed through an ensemble of four different PIMs. Results: The accuracies of the model were 85%, 46%, 69%, 82%, and 93% for KL grade 0, 1, 2, 3, and 4, respectively with the overall accuracy of 76.0%. Conclusions: The ensemble of PIMs could classify knee osteoarthritis severity using simple radiographs with a fine accuracy. Although improvements will be needed in the future, the model has been proven to have the potential to be clinically useful.
참고문헌 (30)
- Kohn MD, "Classifications in brief : Kellgren-Lawrence classification of osteoarthritis", Clin Orthop Relat Res, 474(8), pp.1886~1893 (2016)[DOI]
- Wright RW, "Osteoarthritis Classification Scales : interobserver reliability and arthroscopic correlation", J Bone Joint Surg Am, 96(14), pp.1145~1151 (2014)[DOI]
- Bany Muhammad M, "Interpretable and parameter optimized ensemble model for knee osteoarthritis assessment using radiographs", Sci Rep, 11(1), pp.14348~ (2021)[DOI]
- Thomas KA, "Automated classification of radiographic knee osteoarthritis severity using deep neural networks", Radiol Artif Intell, 2(2), pp.e190065~ (2020)[DOI]
- Liu B, "Toward automatic quantification of knee osteoarthritis severity using improved faster R-CNN", Int J Comput Assist Radiol Surg, 15(3), pp.457~466 (2020)[DOI]
- Tiulpin A, "Automatic grading of individual knee osteoarthritis features in plain radiographs using deep convolutional neural networks", Diagnostics (Basel) (2020)[DOI]
- Abdullah SS, "Automatic detection and classification of knee osteoarthritis using deep learning approach", Radiol Med, 127(4), pp.398~406 (2022)[DOI]
- Swiecicki A, "Deep learning-based algorithm for assessment of knee osteoarthritis severity in radiographs matches performance of radiologists", Comput Biol Med, 133, pp.104334~ (2021)[DOI]
- Kim DH, "Can additional patient information improve the diagnostic performance of deep learning for the interpretation of knee osteoarthritis severity", J Clin Med (2020)[DOI]
- Norman B, "Applying densely connected convolutional neural networks for staging osteoarthritis severity from plain radiographs", J Digit Imaging, 32(3), pp.471~477 (2019)
- "A novel plug-in module for fine-grained visual classification" (2022)[URL]
- Lester G, "The osteoarthritis initiative", Prot Cohort Study, 1, pp.2~ (2006)
- Segal NA, "The Multicenter Osteoarthritis study : opportunities for rehabilitation research", PM R, 5(8), pp.647~654 (2013)[DOI]
- "Damo-yolo: A report on real-time object detection design" (2022)[URL]
- Liu Z, "Swin transformer: hierarchical vision transformer using shifted windows", Proceedings of the IEEE, 2021, 10012~10022
- Tan M, "Efficientnet : rethinking model scaling for convolutional neural networks", International conference on machine learning : PMLR, pp.6105~6114 (2019)
- Chlap P, "A review of medical image data augmentation techniques for deep learning applications", J Med Imaging Radiat Oncol, 65(5), pp.545~563 (2021)
- Zuiderveld K, "Graphics gems", pp.474~485 (1994), Elsevier
- Sane P, "Pixel normalization from numeric data as input to neural networks : For machine learning and image processing", 2017 International Conference on Wireless Communications, Signal Processing and Networking (WiSPNET): IEEE, pp.2221~2225 (2017)
- Szegedy C, "Rethinking the inception architecture for computer vision", Proceedings of the IEEE conference on computer vision and pattern recognition, pp.2818~2826 (2016)
- Kumari S, "An ensemble approach for classification and prediction of diabetes mellitus using soft voting classifier", Int J Cogn Comput Eng, 2, pp.40~46 (2021)
- Muhammad MB, "Eigen-cam : class activation map using principal components", 2020 international joint conference on neural networks (IJCNN): IEEE, pp.1~7 (2020)
- Katharopoulos A, "Stop wasting my FLOPS : improving the efficiency of deep learning models" (2022), EPFL
- Antony J, "Quantifying radiographic knee osteoarthritis severity using deep convolutional neural networks", 2016 23rd international conference on pattern recognition (ICPR): IEEE, pp.1195~1200 (2016)
- Antony J, "O’Connor NE(2017)Automatic detection of knee joints and quantification of knee osteoarthritis severity using convolutional neural networks", Machine learning and data mining in pattern recognition: 13th international conference, MLDM 2017, pp.376~390 (2017), Springer
- Tiulpin A, "Automatic knee osteoarthritis diagnosis from plain radiographs : a deep learning-based approach", Sci Rep, 8(1), pp.1727~ (2018)
- Muhammad MB, "Deep ensemble network for quantification and severity assessment of knee osteoarthritis", 2019 18th IEEE International conference on machine learning and applications (ICMLA): IEEE, pp.951~957 (2019)
- Górriz M, "Assessing knee OA severity with CNN attention-based end-to-end architectures", International conference on medical imaging with deep learning, 102, pp.197~214 (2019), PMLR
- "Automatic grading of knee osteoarthritis on the Kellgren-Lawrence scale from radiographs using convolutional neural networks" (2020)[URL]
- "Fine-grained visual classification with high-temperature refinement and background suppression" (2023)[URL]