한 쌍의 앙상블 모델을 이용한 효율적인 골다공증 예측
Efficient Osteoporosis Prediction Using A Pair of Ensemble Models
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
본 논문에서는 컴퓨터 단층촬영(CT) 이미지를 이용한 합성곱 신경망(CNN)을 기반의 골감소증및 골다공증 예측 모델을 제안한다. 기존의 CNN은 단일 CT 이미지에서 예측에 중요한 지역정보를 활용하지 못하다는 문제가 있다. 본 논문에서 이를 해결하고자 CT 이미지를 정규화하여 질감정보가 다른 두 개의 이미지로 변환하고, 해당 이미지를 활용한 한 쌍의 신경망 네트워크를 제안한다. 동일한 구조를 가진 네트워크 각각의 신경망은 질감 정보가 다른 이미지를 입력으로 사용하고 비유사성 손실함수를 통해 다른 정보를 학습한다. 최종적으로 제안 모델은 중요한 지역정보를 포함한 단일 CT 이미지의 다양한 특징 정보를 학습하며, 이를 앙상블하여 골감소증 및 골다공증 예측 정확도를 높인다. 실험 결과를 통해 제안 모델의 정확도 77.11%를 확인할 수 있으며Grad-CAM을 이용하여 모델이 바라보는 특징을 확인할 수 있다.
Abstract
In this paper, we propose a prediction model for osteopenia and osteoporosis based on a convolutional neural network(CNN) using computed tomography(CT) images. In a single CT image, CNN had a limitation in utilizing important local features for diagnosis. So we propose a compound model which has two identical structures. As an input, two different texture images are used, which are converted from a single normalized CT image. The two networks train different information by using dissimilarity loss function. As a result, our model trains various features in a single CT image which includes important local features, then we ensemble them to improve the accuracy of predicting osteopenia and osteoporosis. In experiment results, our method shows an accuracy of 77.11% and the feature visualize of this model is confirmed by using Grad-CAM.
참고문헌 (19)
- Insha Majeed Wani, "Computer-aided diagnosis systems for osteoporosis detection: a comprehensive survey", Medical & Biological Engineering & Computing, 58(9), pp.1873~1917 (2020), Springer Science and Business Media LLC[DOI]
- Tumay Sozen, "An overview and management of osteoporosis", European Journal of Rheumatology, 4(1), pp.46~56 (2017), AVES Publishing Co.[DOI]
- Andrew D. Smith, "Screening of Bone Density at CT: An Overlooked Opportunity", Radiology, 291(2), pp.368~369 (2019), Radiological Society of North America (RSNA)[DOI]
- Jing Wu, "Healthcare Resource Utilization and Direct Medical Costs for Patients With Osteoporotic Fractures in China", Value in Health Regional Issues, 18, pp.106~111 (2019), Elsevier BV[DOI]
- P. J. Mitchell, "Fracture Liaison Services: the UK experience", Osteoporosis International, 22(S3), pp.487~494 (2011), Springer Science and Business Media LLC[DOI]
- 이재원, "다중 객체검출 기반의 온라인 비디오 요약 알고리즘", 한국컴퓨터정보학회논문지, 24(8), pp.19~28 (2019), 한국컴퓨터정보학회[DOI]
- 황동환, "선택적 커널 합성곱을 이용한 U-Net 기반 망막 혈관 분할 모델", 한국컴퓨터정보학회논문지, 26(4), pp.29~37 (2021), 한국컴퓨터정보학회[DOI]
- Koichiro Yasaka, "Prediction of bone mineral density from computed tomography: application of deep learning with a convolutional neural network", European Radiology, 30(6), pp.3549~3557 (2020), Springer Science and Business Media LLC[DOI]
- Yaling Pan, "Automatic opportunistic osteoporosis screening using low-dose chest computed tomography scans obtained for lung cancer screening", European Radiology, 30(7), pp.4107~4116 (2020), Springer Science and Business Media LLC[DOI]
- Uran Ferizi, "Artificial intelligence, osteoporosis and fragility fractures", Current Opinion in Rheumatology, 31(4), pp.368~375 (2019), Ovid Technologies (Wolters Kluwer Health)[DOI]
- GEIRHOS, Robert, "ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness", arXiv preprint arXiv:1811.12231, ~ (2018)
- ISLAM, Md Amirul, "Shape or texture: Understanding discriminative features in cnns", arXiv preprint arXiv:2101.11604, ~ (2021)
- HE, Kaiming, "Deep residual learning for image recognition", Proceedings of the IEEE conference on computer vision and pattern recognition, pp.770~778 (2016)
- Kim M, "Learning Visual Context by Comparison. In: Vedaldi A., Bischof H., Brox T., Frahm JM. (eds) Computer Vision – ECCV 2020. ECCV 2020", Lecture Notes in Computer Science, 12350 (2020), Springer[DOI]
- DOSOVITSKIY, Alexey, "An image is worth 16x16 words:Transformers for image recognition at scale", arXiv preprint arXiv:2010.11929, ~ (2020)
- TAN, Mingxing, "Efficientnet: Rethinking model scaling for convolutional neural networks", International Conference on Machine Learning. PMLR, 2019, pp.6105~6114 (2019)
- CHOLLET, François, "Xception: Deep learning with depthwise separable convolutions", Proceedings of the IEEE conference on computer vision and pattern recognition, pp.1251~1258 (2017)
- SZEGEDY, Christian, "Rethinking the inception architecture for computer vision", Proceedings of the IEEE conference on computer vision and pattern recognition, pp.2818~2826 (2016)
- SELVARAJU, Ramprasaath R., "Grad-cam: Visual explanations from deep networks via gradient-based localization", Proceedings of the IEEE international conference on computer vision, pp.618~626 (2017)