Abstract
(Aim) Multiple sclerosis is a neurological condition that may cause neurologic disability. Convolutional neural network can achieve good results, but tuning hyperparameters of CNN needs expert knowledge and are difficult and time-consuming. To identify multiple sclerosis more accurately, this article proposed a new transfer-learning-based approach. (Method) DenseNet-121, DenseNet-169, and DenseNet-201 neural networks were compared. In addition, we proposed the use of a composite learning factor (CLF) that assigns different learning factor to three types of layers: early frozen layers, middle layers, and late replaced layers. How to allocate layers into those three layers remains a problem. Hence, four transfer learning settings (viz., Settings A, B, C, and D) were tested and compared. A precomputation method was utilized to reduce the storage burden and accelerate the program. (Results) We observed that DenseNet-201-D (the layers from CP to T3 are frozen, the layers of D4 are updated with learning factor of 1, and the final new layers of FCL are randomly initialized with learning factor of 10) can achieve the best performance. The sensitivity, specificity, and accuracy of DenseNet-201-D was 98.27± 0.58, 98.35± 0.69, and 98.31± 0.53, respectively. (Conclusion) Our method gives better performances than state-of-the-art approaches. Furthermore, this composite learning rate gives superior results to traditional simple learning factor (SLF) strategy.
- P. Browne, D. Chandraratna, C. Angood, H. Tremlett, C. Baker, B. V. Taylor, and A. J. Thompson. 2014. Atlas of multiple sclerosis 2013: A growing global problem with widespread inequity. Neurology 83, 11 (2014), 1022--1024.Google ScholarCross Ref
- J. Benito-leon and M. Laurence. 2019. Malassezia in the central nervous system and multiple sclerosis. Infection 47, 1 (2019), 135--136.Google ScholarCross Ref
- G. Vural, S. Gumusyayla, O. Deniz, S. Neselioglu, and O. Erel. 2019. Relationship between thiol-disulphide homeostasis and visual evoked potentials in patients with multiple sclerosis. Neurological Sciences 40, 2 (2019), 385--391.Google ScholarCross Ref
- A. Didonna. 2016. Preclinical models of multiple sclerosis: Advantages and limitations towards better therapies. Current Medicinal Chemistry 23, 14 (2016), 1442--1459.Google ScholarCross Ref
- E. S. Gromisch, R. D. Kerns, R. Czlapinski, B. Beenken, J. Otis, A. C. Lo, and J. Beauvais. 2019. Cognitive behavioral therapy for the management of multiple sclerosis--related pain: A randomized clinical trial. International Journal of MS Care.Google Scholar
- L. Porten and L. Carrucan-wood. 2017. Caring for a patient with multiple sclerosis. Kai Tiaki: Nursing New Zealand 23, 6 (2017), 16.Google Scholar
- V. Murray, P. Rodriguez, and M. S. Pattichis. 2010. Multiscale AM-FM demodulation and image reconstruction methods with improved accuracy. IEEE Transactions on Image Processing 19, 5 (2010), 1138--1152.Google ScholarDigital Library
- H. Cheng. 2018. Multiple sclerosis identification based on fractional fourier entropy and a modified jaya algorithm. Entropy 20, 4.Google Scholar
- Q. Zhou and X. Shen. 2018. Multiple sclerosis identification by grey-level co-occurrence matrix and biogeography-based optimization. In Proceedings of the 23rd International Conference on Digital Signal Processing (DSP). IEEE, Shanghai, China, 1--5.Google Scholar
- S. Yahia, Y. Ben Salem, and M. N. Abdelkrim. 2018. Texture analysis of magnetic resonance brain images to assess multiple sclerosis lesions. Multimedia Tools and Applications 77, 23 (2018), 30769--30789.Google ScholarDigital Library
- K. J. Xia, J. Q. Wang, and Y. Wu. 2017. Robust alzheimer disease classification based on feature integration fusion model for magnetic. Journal of Medical Imaging and Health Informatics 7, 6 (2017), 1171--1176. https://www.ingentaconnect.com/contentone/asp/jmihi/2017/00000007/00000006/art00008.Google ScholarCross Ref
- Y.-D. Zhang, C. Pan, J. Sun, and C. Tang. 2018. Multiple sclerosis identification by convolutional neural network with dropout and parametric ReLU. Journal of Computational Science 28, 1--10.Google ScholarCross Ref
- A. Sellami and H. Hwang. 2019. A robust deep convolutional neural network with batch-weighted loss for heartbeat classification. Expert Systems with Applications 122, 75--84.Google ScholarDigital Library
- S. Fujino, T. Hatanaka, N. Mori, and K. Matsumoto. 2019. Evolutionary deep learning based on deep convolutional neural network for anime storyboard recognition. Neurocomputing 338, 393--398.Google ScholarDigital Library
- University of Cyprus. 2018. MRI Lesion Segmentation in Multiple Sclerosis Database, eHealth laboratory, University of Cyprus.Google Scholar
- M. Veluchamy and B. Subramani. 2019. Image contrast and color enhancement using adaptive gamma correction and histogram equalization. Optik 183, 329--337.Google ScholarCross Ref
- K. J. Xia, H. S. Yin, and Y. D. Zhang. 2019. Deep semantic segmentation of kidney and space-occupying lesion area based on SCNN and resnet models combined with SIFT-Flow algorithm. Journal of Medical Systems 43, 1 (2019), 12.Google ScholarDigital Library
- K.-J. Xia, H.-S. Yin, and J.-Q. Wang. 2018. A novel improved deep convolutional neural network model for medical image fusion. In Cluster Computing.Google Scholar
- A. Krizhevsky, I. Sutskever, and G. E. Hinton. 2012. Imagenet classification with deep convolutional neural networks. In Proceedings of the 25th International Conference on Neural Information Processing Systems - Volume 1 (Lake Tahoe, Nevada, 2012). Curran Associates Inc., 2999257, 1097--1105.Google Scholar
- A. Z. Karen Simonyan. 2015. Very deep convolutional networks for large-scale image recognition. ICLRGoogle Scholar
- C. Szegedy, L. Wei, J. Yangqing, P. Sermanet, S. Reed, D. Anguelov, D. Erhan, V. Vanhoucke, and A. Rabinovich. 2015. Going deeper with convolutions. In Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 1--9.Google Scholar
- K. He, X. Zhang, S. Ren, and J. Sun. 2016. Deep residual learning for image recognition. In Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), (2016), 9.Google ScholarCross Ref
- G. Huang, Z. Liu, L. V. D. Maaten, and K. Q. Weinberger. 2017. Densely connected convolutional networks. In Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2261--2269.Google Scholar
- X. B. Jin, T. L. Su, J. L. Kong, Y. T. Bai, B. B. Miao, and C. Dou. 2018. State-of-the-art mobile intelligence: Enabling robots to move like humans by estimating mobility with artificial intelligence. Applied Sciences-Basel 8, 3 (2018), 39.Google Scholar
- Y.-D. Zhang, C. Pan, X. Chen, and F. Wang. 2018. Abnormal breast identification by nine-layer convolutional neural network with parametric rectified linear unit and rank-based stochastic pooling. Journal of Computational Science 27, 57--68.Google ScholarCross Ref
- Matlab 2019. Train Deep Learning Network to Classify New Images.Google Scholar
- K. J. Xia, X. W. Zhong, L. Zhang, and J. Q. Wang. 2019. Optimization of diagnosis and treatment of chronic diseases based on association analysis under the background of regional integration. Journal of Medical Systems 43, 3 (2019), 8.Google ScholarDigital Library
- Y. Z. Jiang, K. F. Zhao, K. J. Xia, J. Xue, L. Y. Zhou, Y. Ding, and P. J. Qian. 2019. A novel distributed multitask fuzzy clustering algorithm for automatic MR brain image segmentation. Journal of Medical Systems 43, 5 (2019), 9.Google ScholarDigital Library
- Y. Jin and K. J. Xia. 2018. Intelligent location and recognition mechanism of abnormal point of medical image based on reliable transmission in medial CT scanner local big data networks. Journal of Medical Imaging and Health Informatics 8, 3 609--617.Google ScholarCross Ref
- W. G. Zhu, and C. Chen. 2013. Bionics design of stubble blade based on the digging claw of mole cricket. Applied Mechanics and Materials 397--400, 775--778.Google Scholar
- W. G. Zhu and H. L. Ruan. 2012. Design and research of solar photovoltaic power generation controller based on 89C51 microcontroller. Advanced Materials Research 345, 66--69.Google ScholarCross Ref
- W. Zhu, Q. Zhao, and X. Hu. 2012. Simulation study on hydraulic control system of high-speed transplanter with SIMSAE. In Proceedings of the 2nd International Conference on Consumer Electronics, Communications and Networks (CECNet). IEEE, Yichang, China, 2858--2860.Google Scholar
- F. B. Wang, P. Tu, C. Wu, Y. Liu, and D. Feng. 2018. Femtosecond laser ablation power level identification based on the ablated spot image. International Journal of Advanced Manufacturing Technology 94, 5--8, 2605--2612.Google ScholarCross Ref
- F. B. Wang, A. Xu, K. Zeng, Z. K. Chen, and Y. L. Zhou. 2016. Research for billet limited weight cutting based on behavior stateflow. In Proceedings of the 3rd International Conference on Industrial Engineering and Applications (ICIEA), A. J. Arumugham, M. M. Ulkhaq, M. Kocisko, R. K. Goyal, W. A. Yusmawiza, and X. Qiu, Eds. EDP Sciences, Hong Kong, Peoples R China, Unsp 02005.Google Scholar
- F. B. Wang, C. Wu, Y. Liu, D. Feng, and P. Tu. 2018. Pseudo-color enhancement and its segmentation for femtosecond laser spot image. Microwave and Optical Technology Letters 60, 4 (2018), 854--865.Google ScholarCross Ref
- P. J. Qian, Y. Z. Jiang, Z. H. Deng, L. Z. Hu, S. W. Sun, S. T. Wang, and R. F. Muzic. 2016. Cluster prototypes and fuzzy memberships jointly leveraged cross-domain maximum entropy clustering. IEEE Transactions on Cybernetics 46, 1 (2016), 181--193.Google ScholarCross Ref
- P. J. Qian, K. F. Zhao, Y. Z. Jiang, K. H. Su, Z. H. Deng, S. T. Wang, and R. F. Muzic. 2017. Knowledge-leveraged transfer fuzzy C-Means for texture image segmentation with self-adaptive cluster prototype matching. Knowledge-Based Systems 130, 33--50.Google ScholarCross Ref
- P. J. Qian, Y. Z. Jiang, S. T. Wang, K. H. Su, J. Wang, L. Z. Hu, and R. F. Muzic. 2017. Affinity and penalty jointly constrained spectral clustering with all-compatibility, flexibility, and robustness. IEEE Transactions on Neural Networks and Learning Systems 28, 5 (2017), 1123--1138.Google ScholarCross Ref
- Y. Z. Jiang, Z. H. Deng, F. L. Chung, G. J. Wang, P. J. Qian, K. S. Choi, and S. T. Wang. 2017. Recognition of epileptic EEG signals using a novel multiview TSK fuzzy system. IEEE Transactions on Fuzzy Systems 25, 1 (2017), 3--20.Google ScholarCross Ref
- Y. Z. Jiang, F. L. Chung, S. T. Wang, Z. H. Deng, J. Wang, and P. J. Qian. 2015. Collaborative fuzzy clustering from multiple weighted views. IEEE Transactions on Cybernetics 45, 4 (2015), 688--701.Google ScholarCross Ref
- Y. Z. Jiang, F. L. Chung, H. Ishibuchi, Z. H. Deng, and S. T. Wang. 2015. Multitask TSK fuzzy system modeling by mining intertask common hidden structure. IEEE Transactions on Cybernetics 45, 3 (2015), 548--561.Google ScholarCross Ref
- P. J. Qian, J. X. Zhou, Y. Z. Jiang, F. Liang, K. F. Zhao, S. T. Wang, K. H. Su, and R. F. Muzic. 2018. Multi-view maximum entropy clustering by jointly leveraging inter-view collaborations and intra-view-weighted attributes. IEEE Access 6, 28594--28610.Google ScholarCross Ref
- P. J. Qian, C. Xi, M. Xu, Y. Z. Jiang, K. H. Su, S. T. Wang, and R. F. Muzic. 2018. SSC-EKE: Semi-supervised classification with extensive knowledge exploitation. Information Sciences 422, 51--76.Google ScholarDigital Library
- P. J. Qian, S. W. Sun, Y. Z. Jiang, K. H. Su, T. G. Ni, S. T. Wang, and R. F. Muzic. 2016. Cross-domain, soft-partition clustering with diversity measure and knowledge reference. Pattern Recognition 50, 155--177.Google ScholarDigital Library
Index Terms
- DenseNet-201-Based Deep Neural Network with Composite Learning Factor and Precomputation for Multiple Sclerosis Classification
Recommendations
Deep learning based Self-Sustained Personal Network
AbstractMost of the research on personal networking and deep learning have been conducted separately. Crossovers between the two fields have just emerged. This article provides a quick introduction to the fundamentals of deep learning, as well as the most ...
A Study on Improved Deep Learning Structure Based on DenseNet
The existing image-related deep learning research methods are conducted through algorithms based on feature identification and association, but there are limits to their accuracy and reliability. These methods are inefficient for artificial neural ...
Deep neural network with transfer learning in remote object detection from drone
DroneCom '22: Proceedings of the 5th International ACM Mobicom Workshop on Drone Assisted Wireless Communications for 5G and BeyondIn this article we present a model of new deep learning composition for remote ship detection. Proposed architecture is composed of newly developed derivatives of ResNet, DenseNet and CNN composed into one global classifier. Since training of such model ...
Comments