Real-Time Noise Classification of Medical Image via Online Machine Learning Algorithm
AbstractMedical image is generally deteriorated by noise due to signal acquisition, signal processing, and other reasons. Noise classification of medical image is able to enhance post-processing tasks like medical image segmentation, registration, and analysis. Due to real-time requirements of medical image analysis for clinical applications, noise classification of medical image is desired to be fast for meeting real-time requirements. On the other hand, online learning algorithms have been studied for processing online data in real-time mode, which can produce rapid learning model based on adjustments of new incoming data. In this paper, we investigate perceptron algorithm - a classical online learning method for noise classification in parallel magnetic resonance imaging (pMRI). Noise generated in pMRI is quickly classified and online classification model is updated in real-time simultaneously. Experimental results demonstrate that noise and brain tissues existing MR images is able to be classified dynamically with the perceptron algorithm.
(1) Gondara, L, Medical image denoising using convolutional denoising autoencoders. Data Mining Workshops (ICDMW), IEEE 16th International Conference on 2016.
(2) Kohan, M.N., et al., Denoising medical images using calculus of variations. Journal of Medical Signals and Sensors, 2011. 1(3): p. 184-190.
(3) Seetha J., et al., Denoising of MRI images using filtering methods. Wireless Communications, Signal Processing and Networking (WiSPNET), International Conference on 2016.
(4) Shen, D., Deep learning in medical image analysis. Annual Review of Biomedical Engineering, 2017. 19: p. 221-248.
(5) Zhou, S.k., Medical image recognition, segmentation and parsing: machine learning and multiple object approaches. The Elsevier and Miccai Society Book Series, Academic Press 2016.
(6) Guido, T., et al., Model attraction in medical image object recognition. Proceedings of the SPIE, 1995. 2436(1): p. 18-29.
(7) Olesen O.V., et al., Motion tracking for medical imaging: a nonvisible structured light tracking approach. Medical Imaging, IEEE Transactions on, 2012. 31(1): p. 79-87.
(8) Lim J.H., et al., Motion tracking in medical images. Biomedical Image Understanding, Methods and Applications, 2015.
(9) Huang C., et at., Real-time 3D motion tracking using MR micro-coils
for PET imaging. The Journal of Nuclear Medicine, 2013. 54(2): 44.
(10) Cleary, K., et al., Image-guided interventions: technology review and clinical applications. Annual Review of Biomedical Engineering, 2010. 12: p. 119-142.
(11) Shalev-Shwartz, S., Online learning and online convex optimization. Foundations and Trends in Machine Learning, 2011. 4(12: p. 107-194.
(12) Hoi, S.C.H, LIBOL: a library for online learning algorithms. Journal of Machine Learning Research, 2014. 12(1): p. 495-499.
(13) Pruessmann, K.P., et al., SENSE: sensitivity encoding for fast MRI. Magnetic Resonance in Medicine, 1999. 45(2): p. 952-962.
(14) URL: http://www.mr-tip.com/serv1.php?type=cam.
(15) Rosenblatt, F., The perceptron: a probabilistic model for information storage and organization in the brain. Psychological Review, 1958. 65(6): p. 386-408.
(16) Cesa-Bianchi, N., et al. A second-order perceptron algorithm. SIAM Journal on Computing, 2005. 34(3): 640-668.
(17) Pieciak, T., et al., Non-stationary rician noise estimation in parallel MRI using a single image: a variance-stabilizing approach. Pattern Analysis and Machine Intelligence, IEEE Transactions on, 2017. 39(10): p. 2015-2029.
Aja-Fernandez, S., et al., Statistical noise analysis in GRAPPA using a parametrized noncentral Chi approximation model. Magnetic Resonance in Medicine, 2011. 65(4): p. 1195-1206.