在臨床上，白質高訊號區被認為是腦部血管疾病的重要特徵。為了研究白質高訊號區跟腦部血管疾病的關聯性，我們需要先分割出白質高訊號區。 我們使用白質高訊號區較容易辨別的T2權重液體衰減核磁共振影像作為我們的研究材料。因為掃描出來的影像會有亮度的差異，為了讓我們的方法也適用在亮度不同的影像上，我們統計個別影像濾除背景後出現像素值的眾數，使用分段線性函數及珈瑪校正將其調整為數據集裡的所有影像中濾除背景後出現最多次的像素值，並以調整完的影像作為訓練資料。我們利用卷積神經網路中的U-net架構作為我們分割的工具。我們將U-net的分割結果與現有可以分割白質高訊號區的病灶分割工具中的病變預測演算法做比較，發現我們的方法能將白質高訊號區分割得更準確。

In clinical practice, white matter hyperintensities(WMH) are considered an important feature of cerebral vascular disease. To study the relationship between WMH and cerebral vascular disease, it is necessary to first segment the WMH. We used T2-weighted fluid-attenuated inversion recovery (T2-FLAIR) magnetic resonance imaging(MRI) as our research material, since WMH features are distinct in these images. Since the produced images may vary in brightness, to make our method applicable to images with different brightness levels, we calculated the most frequently occurring pixel value in each image after background removal. We then used a piecewise linear transformation and gamma correction to adjust it to match the most frequently occurring pixel value across all images in the dataset after background removal, and used these adjusted images as the training data. We utilized U-net framework which is an architecture of convolution neural networks as our segmentation tool. We compared the segmentation results by U-net framework with those from the lesion prediction algorithm(LPA) within the lesion segmentation tool(LST) which is a toolbox that can be used in segmenting WMH, and found that our method can help us to segment WMH more accurately.

[1] Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation. ArXiv, abs/1505.04597, 2015.
[2] Luca Melazzini, Clare E. Mackay, Valentina Bordin, Sana Suri, Enikő Zsoldos, Nicola Filippini, Abda Mahmood, Vaanathi Sundaresan, Marina Codari, Eugene Duff, Archana Singh-Manoux, Mika Kivimäki, Klaus P. Ebmeier, Mark Jenkinson, Francesco Sardanelli, and Ludovica Griffanti. White matter hyperintensities classified according to intensity and spatial location reveal specific associations with cognitive performance. NeuroImage: Clinical, 30:102616, 2021.
[3] Markus D. Schirmer, Adrian V. Dalca, Ramesh Sridharan, Anne-Katrin Giese, Kathleen L. Donahue, Marco J. Nardin, Steven J.T. Mocking, Elissa C. McIntosh, Petrea Frid, Johan Wasselius, John W. Cole, Lukas Holmegaard, Christina Jern, Jordi Jimenez-Conde, Robin Lemmens, Arne G. Lindgren, James F. Meschia, Jaume Roquer, Tatjana Rundek, Ralph L. Sacco, Reinhold Schmidt, Pankaj Sharma, Agnieszka Slowik, Vincent Thijs, Daniel Woo, Achala Vagal, Huichun Xu,Steven J. Kittner, Patrick F. McArdle, Braxton D. Mitchell, Jonathan Rosand, Bradford B. Worrall, Ona Wu, Polina Golland, and Natalia S. Rost. White matter hyperintensity quantification in large-scale clinical acute ischemic stroke cohorts –the mri-genie study. NeuroImage: Clinical, 23:101884, 2019.
[4] Elisa R. Torres, Emily F. Strack, Claire E. Fernandez, Tyler A. Tumey, and Mary E. Hitchcock. Physical activity and white matter hyperintensities: A systematic review of quantitative studies. Preventive Medicine Reports, 2:319–325, 2015.
[5] J Botz, V Lohner, and MD Schirmer. Spatial patterns of white matter hyperintensities: a systematic review. Frontiers in Aging Neuroscience, 15:1165324, 2023.
[6] V Farkhondeh and C DeCarli. White matter hyperintensities in diverse populations: A systematic review of literature in the united states. Cerebral Circulation - Cognition and Behavior, 6:100204, 2024.
[7] Paul Schmidt and Lucie Wink. LST: A lesion segmentation tool for SPM. Manual/Documentation for version 3.0.0, October 2019. Version 3.0.0, Technische Universität München, Munich, Germany.
[8] Paul Schmidt, Christian Gaser, Milan Arsic, Dorothea Buck, Annette Förschler, Achim Berthele, Muna Hoshi, Rüdiger Ilg, Volker J. Schmid, Claus Zimmer, Bernhard Hemmer, and Mark Mühlau. An automated tool for detection of flairhyperintense white-matter lesions in multiple sclerosis. NeuroImage, 59(4):3774– 3783, 2012.
[9] Paul Schmidt. Bayesian inference for structured additive regression models for large-scale problems with applications to medical imaging, January 2017.
[10] Hugo J. Kuijf, J. Matthijs Biesbroek, Jeroen De Bresser, Rutger Heinen, Simon Andermatt, Mariana Bento, Matt Berseth, Mikhail Belyaev, M. Jorge Cardoso, Adrià Casamitjana, D. Louis Collins, Mahsa Dadar, Achilleas Georgiou, Mohsen Ghafoorian, Dakai Jin, April Khademi, Jesse Knight, Hongwei Li, Xavier Lladó, Miguel Luna, Qaiser Mahmood, Richard McKinley, Alireza Mehrtash, Sébastien Ourselin, Bo-Yong Park, Hyunjin Park, Sang Hyun Park, Simon Pezold, Elodie Puybareau, Leticia Rittner, Carole H. Sudre, Sergi Valverde, Verónica Vilaplana, Roland Wiest, Yongchao Xu, Ziyue Xu, Guodong Zeng, Jianguo Zhang, Guoyan Zheng, Christopher Chen, Wiesje van der Flier, Frederik Barkhof, Max A. Viergever, and Geert Jan Biessels. Standardized assessment of automatic segmentation of white matter hyperintensities and results of the wmh segmentation challenge. IEEE Transactions on Medical Imaging, 38(11):2556–2568, 2019.
[11] Y. LeCun, B. Boser, J. S. Denker, D. Henderson, R. E. Howard, W. Hubbard, and L. D. Jackel. Backpropagation applied to handwritten zip code recognition. Neural Computation, 1(4):541–551, 1989.
[12] Y. Lecun, L. Bottou, Y. Bengio, and P. Haffner. Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11):2278–2324, 1998.
[13] Murilo Gustineli. A survey on recently proposed activation functions for deep learning. ArXiv, abs/2204.02921, 2022.
[14] Alex Krizhevsky, Ilya Sutskever, and Geoffrey E. Hinton. Imagenet classification with deep convolutional neural networks. In F. Pereira, C. J. C. Burges, L. Bottou, and K. Q. Weinberger, editors, Advances in Neural Information Processing Systems 25, pages 1097–1105. Curran Associates, Inc., 2012.
[15] Bing Xu, Naiyan Wang, Tianqi Chen, and Mu Li. Empirical evaluation of rectified activations in convolutional network. CoRR, abs/1505.00853, 2015.
[16] Juan R. Terven, Diana-Margarita Córdova-Esparza, Alfonso Ramírez-Pedraza, and Edgar A. Chavez-Urbiola. Loss functions and metrics in deep learning. 2023.
[17] Sebastian Ruder. An overview of gradient descent optimization algorithms. ArXiv, abs/1609.04747, 2016.
[18] Boris Polyak. Some methods of speeding up the convergence of iteration methods. Ussr Computational Mathematics and Mathematical Physics, 4:1–17, 12 1964.
[19] Tijmen Tieleman and Geoffrey Hinton. Lecture 6.5-rmsprop: Divide the gradient by a running average of its recent magnitude. COURSERA: Neural networks for machine learning, 4(2):26–31, 2012.
[20] Diederik P. Kingma and Jimmy Ba. Adam: A method for stochastic optimization. CoRR, abs/1412.6980, 2014.
[21] Sergey Ioffe and Christian Szegedy. Batch normalization: Accelerating deep network training by reducing internal covariate shift. ArXiv, abs/1502.03167, 2015.
[22] Brittany Reiche, A.R. Moody, and April Khademi. Pathology-preserving intensity standardization framework for multi-institutional flair mri datasets. Magnetic Resonance Imaging, 62:59–69, 2019.
[23] Lucia Torres-Simón, Alberto del Cerro-León, Miguel Yus, Ricardo Bruña, Lidia Gil-Martínez, Alberto Marcos Dolado, Fernando Maestú, Juan Arrazola-García, and Pablo Cuesta. Comparison between tools for automatic segmentation of white matter hyperintensities of presumed vascular origin in aging: which one, how and why to choose the most suitable for your purpose? medRxiv : the preprint server for health sciences, 2023.
[24] Quentin Vanderbecq, Eric Xu, Sebastian Ströer, Baptiste Couvy-Duchesne, Mauricio Diaz Melo, Didier Dormont, and Olivier Colliot. Comparison and validation of seven white matter hyperintensities segmentation software in elderly patients. NeuroImage: Clinical, 27:102357, 2020.
[25] Lauren M Hablitz, Virginia Plá, Marta Giannetto, Hilaire S Vinitsky, Filippe F Stæger, Timothy Metcalfe, Robert Nguyen, Anouar Benrais, and Maiken Nedergaard. Circadian control of brain glymphatic and lymphatic fluid flow. Nat Commun, 11(1):4411, Sep 2020.
[26] Lauren M Hablitz and Maiken Nedergaard. The glymphatic system: A novel component of fundamental neurobiology. J Neurosci, 41(37):7698–7711, Sep 2021.
[27] Toshiaki Taoka, Yoshitaka Masutani, Hisashi Kawai, Toshiki Nakane, Kei Matsuoka, Fumihiko Yasuno, Toshifumi Kishimoto, and Shinji Naganawa. Evaluation of glymphatic system activity with the diffusion MR technique: diffusion tensor image analysis along the perivascular space (DTI-ALPS) in alzheimer’s disease cases. Jpn J Radiol, 35(4):172–178, Apr 2017.
[28] Wei Zhang, Yuyan Zhou, Jing Wang, Xiaokang Gong, Zhongping Chen, Xiaohuan Zhang, Jing Cai, Shengdi Chen, Li Fang, Jie Sun, and Min Lou. Glymphatic clearance function in patients with cerebral small vessel disease. Neuroimage, 238:118257, Sep 2021.