秀麗隱桿線蟲是一種多細胞生物，與人類基因有百分之六十的相似性。由於線蟲體型小、透明、生命週期短、飼養繁殖容易以及基因完全定序，常提供為多項研究的動物模型。1986年，科學家已繪製出線蟲神經系統中每一個神經元的位置。線蟲利用有限的神經元，可感知周圍環境變化立刻產生相對應之運動行為，因此適合用於探討神經傳導與行為輸出和基因的關聯性。若能有更好比較與量化線蟲行為模式的方法，將有助於神經科學、基因工程甚至環境毒物學等多重領域應用。
線蟲在不同環境的前行方式有兩種，在液體中游動與在濃稠的凝膠上爬行。本研究利用近年發展的適應性數據分析方法希爾伯特-黃轉換對野生株與多種基因突變株線蟲進行步態分析。透過希爾伯特-黃轉換的兩大步驟，經驗模態分解和希爾伯特頻譜分析，可以分析非穩態與非線性訊號。近來研究顯示，由經驗模態分解得出的固有模態函數可能含有實際的物理意義。
此研究使用暗視野顯微鏡觀測線蟲在含有生長培養基的聚二甲基矽氧烷填充槽內之運動行為，並利用高速攝影機拍攝。以線蟲身體中央線隨時間變化的曲率圖作為運動特徵，透過經驗模態分解得出的固有模態函數，分別與不同品種的線蟲進行相關性比較，即可得到線蟲在身體各部位的運動特徵作為線蟲步態指紋。由結果顯示，在影像上看起來同樣為蛇行或滾動式爬行的線蟲，經由計算它們的步態指紋仍有差異性，可透過非監督式學習的演算法分至不同的群集。
此研究的新穎之處在於可以藉由比較不同線蟲的步態指紋來分類不同品種的線蟲以及計算出每隻線蟲與其他品種線蟲的運動行為相似度。未來也可以應用於量化線蟲身體受損或發病後所造成運動行為的改變。

The round worm Caenorhabditis (C.) elegans is used as a multicellular animal model for a wide variety of research and shares 60% genetic similarities with the human being. The advantages of C. elegans include the small size, transparency, short life cycles, ease of cultivation and fully-sequenced genome. In 1986, scientists have mapped all the neurons in its nerve systems. With 302 neurons, C. elegans can sense its surrounding environment and respond by changing its locomotion. Therefore, it is an excellent model for research of neural circuits, behavior outputs and the genetic relationships. Developing useful methods to compare and quantify the locomotion of the worms can be useful in multiple research fields, such as neuroscience, genetic engineering and even environmental toxicology.
C. elegans has two ways of locomotion: swimming in liquid and crawling on dense gels. In this study, we use a wild-type strain (N2) and several mutant strains (CB0061, TJ356, CL2070, CL2120) and apply the newly developed adaptive data analysis method Hilbert-Huang Transform (HHT) in the gait analysis. HHT consists of two parts, empirical mode decomposition (EMD) and Hilbert spectral analysis (HSA) for analyzing nonstationary and nonlinear signals. Recent studies show that the intrinsic mode functions (IMFs) derived by EMD may carry actual physical significance.
The locomotion of C. elegans in a chamber of Polydimethylsiloxane (PDMS) filled with nematode growth media (NGM) was recorded with a CCD camera under a dark field microscope. Kymograph is used as the characteristics of the worm’s locomotion which is comprised of the curvature curves of the central line of the worm body varying with time. By using multidimensional empirical mode decomposition (MEMD), kymographs can be decomposed into several IMFs. Comparing the IMFs at different body parts of a worm using correlation algorithm, we can obtained its gaitprint. The results show that similar gaits in the same classification such as sinusoidal wave and roller still differ from one another in the gaitprint. Unsupervised learning method can be used for clustering the worms into several groups.
The novelty of the research is that we create automated gaitprint analysis for the comparison of different gaits and provide a resemblance score for identifying different strains. The method in this study can be further applied to find out new gaits, physical defects and gait changes due to diseases of C. elegans.

[1] P. Sengupta and A. D. T. Samuel, “C. elegans: a model system for systems neuroscience,” Curr. Opin. Neurobiol., vol. 19, no. 6, pp. 637–643, 2009.
[2] S. Brenner, “The genetics of Caenorhabditis elegans,” Genetics, vol. 77, no. 1, pp. 71–94, 1974.
[3] H. E. Kinser and Z. Pincus, “High-throughput screening in the C. elegans nervous system,” Mol. Cell. Neurosci., 2016.
[4] D. J. Dickinson and B. Goldstein, “CRISPR-Based Methods for Caenorhabditis elegans Genome Engineering,” Genetics, vol. 202, no. 3, pp. 885–901, 2016.
[5] D. Zhou, J. Yang, H. Li, C. Cui, Y. Yu, Y. Liu, and K. Lin, “The chronic toxicity of bisphenol A to Caenorhabditis elegans after long-term exposure at environmentally relevant concentrations,” Chemosphere, vol. 154, pp. 546–551, 2016.
[6] D. N. Lee and H. Kalmus, “The structure of the nervous system of the nematode Caenorhabditis elegans,” Philos. Trans. R. Soc. London Biol. Sci., vol. 314, no. 1165, pp. 1–340, 1986.
[7] The C. elegans Sequencing Consortium, “Genome sequence of the nematode C. elegans: a platform for investigating biology,” Science, vol. 282, pp. 2012–2018, 1998.
[8] K. W. Jordan, M. A. Carbone, A. Yamamoto, T. J. Morgan, and T. F. C. Mackay, “Quantitative genomics of locomotor behavior in Drosophila melanogaster,” Genome Biol., vol. 8, no. 8, pp. 1–17, 2007.
[9] B. J. Piggott, J. Liu, Z. Feng, S. A. Wescott, and X. Z. S. Xu, “The neural circuits and synaptic mechanisms underlying motor initiation in C. elegans,” Cell, vol. 147, no. 4, pp. 922–933, 2011.
[10] M. Dimitriadi and A. C. Hart, “Neurodegenerative disorders: insights from the nematode Caenorhabditis elegans,” Neurobiol. Dis., vol. 40, no. 1, pp. 4–11, 2010.
[11] A. E. X. Brown and W. R. Schafer, “Automated behavioural fingerprinting of C. elegans mutants,” arXiv, vol. 1301, no. 1017, pp. 1–28, 2013.
[12] A. Vidal-Gadea, S. Topper, L. Young, A. Crisp, L. Kressin, E. Elbel, T. Maples, M. Brauner, K. Erbguth, A. Axelrod, A. Gottschalk, D. Siegel, and J. T. Pierce-Shimomura, “Caenorhabditis elegans selects distinct crawling and swimming gaits via dopamine and serotonin,” Proc. Natl. Acad. Sci., vol. 108, no. 42, pp. 17504–17509, 2011.
[13] V. J. Butler, R. Branicky, E. Yemini, J. F. Liewald, A. Gottschalk, R. A. Kerr, D. B. Chklovskii, and W. R. Schafer, “A consistent muscle activation strategy underlies crawling and swimming in Caenorhabditis elegans,” J. R. Soc. Interface, vol. 12, no. 102, pp. 1–12, 2015.
[14] J. Hodgkin, “Male phenotypes and mating efficiency in Caenorhabditis elegans,” Genetics, vol. 103, no. 1, pp. 43–64, 1983.
[15] I. L. Johnstone, “Cuticle collagen genes: expression in Caenorhabditis elegans,” Trends Genet., vol. 16, no. 1, pp. 21–27, 2000.
[16] S. J. Husson, W. S. Costa, C. Schmitt, and A. Gottschalk, “Keeping track of worm trackers,” in WormBook, 2012, pp. 1–17.
[17] J.-H. Baek, P. Cosman, Z. Feng, J. Silver, and W. R. Schafer, “Using machine vision to analyze and classify Caenorhabditis elegans behavioral phenotypes quantitatively,” J. Neurosci. Methods, vol. 118, no. 1, pp. 9–21, 2002.
[18] C. Restif, C. Ibáñez-Ventoso, M. M. Vora, S. Guo, D. Metaxas, and M. Driscoll, “CeleST: computer vision software for quantitative analysis of C. elegans swim behavior reveals novel features of locomotion,” PLoS Comput. Biol., vol. 10, no. 7, pp. 1–12, 2014.
[19] J. H. Boyle, S. Berri, and N. Cohen, “Gait modulation in C. elegans: an integrated neuromechanical model,” Front. Comput. Neurosci., vol. 6, no. 10, pp. 1–15, 2012.
[20] E. Yemini, T. Jucikas, L. J. Grundy, A. E. Brown, and W. R. Schafer, “A database of Caenorhabditis elegans behavioral phenotypes,” Nat. Methods, vol. 10, no. 9, pp. 877–879, 2013.
[21] W.-J. Kuo, Y.-S. Sie, and H.-S. Chuang, “Characterizations of kinetic power and propulsion of the nematode Caenorhabditis elegans based on a micro-particle image velocimetry system,” Biomicrofluidics, vol. 8, no. 2, pp. 1–13, 2014.
[22] K.-M. Huang, P. Cosman, and W. R. Schafer, “Machine vision based detection of omega bends and reversals in C. elegans,” J. Neurosci. Methods, vol. 158, no. 2, pp. 323–336, 2006.
[23] G. J. Stephens, B. Johnson-Kerner, W. Bialek, and W. S. Ryu, “Dimensionality and dynamics in the behavior of C. elegans,” PLoS Comput. Biol., vol. 4, no. 4, pp. 1–10, 2008.
[24] A. E. X. Brown, E. I. Yemini, L. J. Grundy, T. Jucikas, and W. R. Schafer, “A dictionary of behavioral motifs reveals clusters of genes affecting Caenorhabditis elegans locomotion,” Proc. Natl. Acad. Sci., vol. 110, no. 2, pp. 791–796, 2013.
[25] Y. Koren, R. Sznitman, P. E. Arratia, C. Carls, P. Krajacic, A. E. X. Brown, and J. Sznitman, “Model-independent phenotyping of C. elegans locomotion using scale-invariant feature transform,” PLoS One, vol. 10, no. 3, pp. 1–16, 2015.
[26] D. G. Lowe, “Distinctive image features from scale invariant keypoints,” Int. J. Comput. Vis., vol. 60, no. 2, pp. 91–110, 2004.
[27] N. E. Huang, Z. Shen, S. R. Long, M. C. Wu, H. H. Shih, Q. Zheng, N.-C. Yen, C. C. Tung, and H. H. Liu, “The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis,” Proc. R. Soc. London A Math. Phys. Eng. Sci., vol. 454, no. 1971, pp. 903–995, 1998.
[28] N. E. Huang and Z. Wu, “A review on Hilbert-Huang Transform: method and its applications to geophysical studies,” Rev. Geophys., vol. 46, no. 2, pp. 1–23, 2008.
[29] N. E. Huang, M.-L. Wu, W. Qu, S. R. Long, and S. S. P. Shen, “Applications of Hilbert-Huang transform to non-stationary financial time series analysis,” Appl. Stoch. Model. Bus. Ind., vol. 19, no. 3, pp. 245–268, 2003.
[30] L. Xiang and A. Hu, “Comparison of methods for different time-frequency analysis of vibration signal,” J. Softw., vol. 7, no. 1, pp. 68–74, 2012.
[31] P. Kuchi and S. Panchanathan, “Intrinsic mode functions for gait recognition,” Proc. 2004 Int. Symp. Circuits Syst., vol. II, pp. 117–120, 2004.
[32] O. Sahu, V. Anand, V. Kanhangad, and R. B. Pachori, “Classification of magnetic resonance brain images using bi-dimensional empirical mode decomposition and autoregressive model,” Biomed. Eng. Lett., vol. 5, no. 4, pp. 311–320, 2015.
[33] G. Wang, X.-Y. Chen, and F.-L. Qiao, “On intrinsic mode function,” Adv. Adapt. Data Anal., vol. 2, no. 3, pp. 277–293, 2010.
[34] Z. Wu, N. E. Huang, and X. Chen, “The multi-dimensional ensemble empirical mode decomposition method,” Adv. Adapt. Data Anal., vol. 1, no. 3, pp. 339–372, 2009.
[35] J. M. Kramer, R. P. French, E.-C. Park, and J. J. Johnson, “The Caenorhabditis elegans rol-6 gene, which interacts with the sqt-1 collagen gene to determine organismal morphology, encodes a collagen,” Mol. Cell. Biol., vol. 10, no. 5, pp. 2081–2089, 1990.
[36] J. M. Kramer and J. J. Johnson, “Analysis of mutations in the sqt-1 and rol-6 collagen genes of Caenorhabditis elegans,” Genetics, vol. 135, no. 4, pp. 1035–1045, 1993.
[37] C. Thacker, J. A. Sheps, and A. M. Rose, “Caenorhabditis elegans dpy-5 is a cuticle procollagen processed by a proprotein convertase,” Cell. Mol. Life Sci., vol. 63, no. 10, pp. 1193–1204, 2006.
[38] G. McColl, B. R. Roberts, A. P. Gunn, K. A. Perez, D. J. Tew, C. L. Masters, K. J. Barnham, R. A. Cherny, and A. I. Bush, “The Caenorhabditis elegans Aβ1-42 model of Alzheimer disease predominantly expresses Aβ3-42,” J. Biol. Chem., vol. 284, no. 34, pp. 22697–22702, 2009.
[39] M. Porta-de-la-Riva, L. Fontrodona, A. Villanueva, and J. Cerón, “Basic Caenorhabditis elegans methods: synchronization and observation,” J. Vis. Exp., no. 64, pp. 1–9, 2012.
[40] Z. F. Altun and D. H. Hall, “Introduction to C. elegans Anatomy,” in In WormAtlas, 2005.
[41] J. C. Mcdonald, D. C. Duffy, J. R. Anderson, D. T. Chiu, H. Wu, O. J. A. Schueller, and G. M. Whitesides, “Fabrication of microfluidic systems in poly (dimethylsiloxane),” Electrophoresis, vol. 21, no. 1, pp. 27–40, 2000.
[42] J. Canny, “A computational approach to edge detection,” IEEE Trans. Pattern Anal. Mach. Intell., vol. PAMI-8, no. 6, pp. 679–698, 1986.
[43] Y.-H. Wang, C.-H. Yeh, H.-W. V. Young, K. Hu, and M.-T. Lo, “On the computational complexity of the empirical mode decomposition algorithm,” Phys. A Stat. Mech. its Appl., vol. 400, pp. 159–167, 2014.
[44] P. J. Rousseeuw, “Silhouettes: a graphical aid to the interpretation and validation of cluster analysis,” J. Comput. Appl. Math., vol. 20, pp. 53–65, 1987.
[45] G. R. Francis and R. H. Waterston, “Muscle organization in Caenorhabditis elegans: localization of proteins implicated in thin filament attachment and I-band organization,” J. Cell Biol., vol. 101, no. 4, pp. 1532–1549, 1985.
[46] E. B. Maryon, R. Coronado, and P. Anderson, “unc-68 encodes a ryanodine receptor involved in regulating C. elegans body-wall muscle contraction,” J. Cell Biol., vol. 134, no. 4, pp. 885–893, 1996.
[47] J. Karbowski, C. J. Cronin, A. Seah, J. E. Mendel, D. Cleary, and P. W. Sternberg, “Conservation rules, their breakdown, and optimality in Caenorhabditis sinusoidal locomotion,” J. Theor. Biol., vol. 242, no. 3, pp. 652–669, 2006.