Dendritic tree extraction from noisy maximum intensity projection images in C. elegans

BioMedical Engineering OnLine

Ayala Greenblum³  Raphael Sznitman⁴  Pascal Fua⁴  Paulo E Arratia²  Meital Oren¹  Benjamin Podbilewicz¹  Josué Sznitman³ 

[1] Current address: Department of Biochemistry & Molecular Biophysics, Columbia University, 1032, New York, USA

[2] Department of Mechanical Engineering and Applied Mechanics, University of Pennsylvania, 19104, Philadelphia, USA

[3] Department of Biomedical Engineering, Technion - Israel Institute of Technology, 32000, Haifa, Israel

[4] School of Computer and Communications, Ecole Polytechnique Fédérale de Lausanne (EPFL), 1015, Lausanne, Switzerland

关键词: Neuronal arborization;    Bayesian probability;    Statistical learning;    Image segmentation;    Computer vision;    C. elegans;    Neuronal dendrites;   

Others  :  1084837
DOI  :  10.1186/1475-925X-13-74

 received in 2013-11-09, accepted in 2014-05-27,  发布年份 2014

【 摘 要 】

Background

Maximum Intensity Projections (MIP) of neuronal dendritic trees obtained from confocal microscopy are frequently used to study the relationship between tree morphology and mechanosensory function in the model organism C. elegans. Extracting dendritic trees from noisy images remains however a strenuous process that has traditionally relied on manual approaches. Here, we focus on automated and reliable 2D segmentations of dendritic trees following a statistical learning framework.

Methods

Our dendritic tree extraction (DTE) method uses small amounts of labelled training data on MIPs to learn noise models of texture-based features from the responses of tree structures and image background. Our strategy lies in evaluating statistical models of noise that account for both the variability generated from the imaging process and from the aggregation of information in the MIP images. These noisy models are then used within a probabilistic, or Bayesian framework to provide a coarse 2D dendritic tree segmentation. Finally, some post-processing is applied to refine the segmentations and provide skeletonized trees using a morphological thinning process.

Results

Following a Leave-One-Out Cross Validation (LOOCV) method for an MIP databse with available “ground truth” images, we demonstrate that our approach provides significant improvements in tree-structure segmentations over traditional intensity-based methods. Improvements for MIPs under various imaging conditions are both qualitative and quantitative, as measured from Receiver Operator Characteristic (ROC) curves and the yield and error rates in the final segmentations. In a final step, we demonstrate our DTE approach on previously unseen MIP samples including the extraction of skeletonized structures, and compare our method to a state-of-the art dendritic tree tracing software.

Conclusions

Overall, our DTE method allows for robust dendritic tree segmentations in noisy MIPs, outperforming traditional intensity-based methods. Such approach provides a useable segmentation framework, ultimately delivering a speed-up for dendritic tree identification on the user end and a reliable first step towards further morphological characterizations of tree arborization.

【 授权许可】

   
2014 Greenblum et al.; licensee BioMed Central Ltd.

【 图 表 】

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

【 参考文献 】

• [1]Tan P, Katsanis N: Thermosensory and mechanosensory perception in human genetic disease. Hum Mol Genet 2009, 18:146-155.
• [2]Scott E, Luo L: How do dendrites takes their shape? Nat Neurosci 2001, 4:359-365.
• [3]Syntichaki P, Tavernarakis N: Genetic models of mechanotransduction: the nematode Caenorhabditis elegans. Physiol Rev 2004, 84:1097-1153.
• [4]Yan Y, Yan L: Branching out: mechanisms of dendritic arborization. Nat Rev Neurosci 2010, 11:316-328.
• [5]Albeg A, Smith C, Chatzigeorgiou M, Feitelson D, Hall D, Schafer W, Miller D, Treinin M: C. elegans multi-dendritic sensory neurons: morphology and function. Mol Cell Neu 2011, 46:308-317.
• [6]Oren-Suissa M, Hall D, Treinin M, Shemer G, Podbilewicz B: The fusogen EFF-1 controls sculpting of mechanosensory dendrites. Science 2010, 328:1285-1288.
• [7]Brenner S: The genetics of Caenorhabditis elegans. Genetics 1974, 77:71-94.
• [8]Consortium CES: Genome sequence of the nematode C. elegans: a platform for investigating biolog. Science 1998, 282:2012-2018.
• [9]Sulston JE, Schierenberg E, White JG, Thomson JN: The embryonic cell lineage of the nematode Caenorhabditis elegans. Dev Biol 1983, 100:64-119.
• [10]Nam SW, Qian C, Kim S, van Noort D, Chiam KH, Park S: C. elegans sensing of and entrainment along obstacles require different neurons at different body locations. Sci Rep 2013, 3:3247.
• [11]White JG, Southgate E, Thomson JN, Brenner S: The structure of the ventral nerve cord of Caenorhabditis elegans. Phil Trans R Soc Lond B Biol Sci 1976, 275:327-348.
• [12]White JG, Southgate E, Thomson JN, Brenner S: The structure of the nervous system of the nematode C. elegans. Phil Trans R Soc Lond B Biol Sci 1986, 314:1-340.
• [13]Lumpkin E, Catarina M: Mechanisms of sensory transduction in the skin. Nature 2007, 445:858-865.
• [14]Aguirre-Chen C, Bullow H, Kaprielian Z: C. elegans bicd-1, homolog of the Drosophila dynein accessory factor Bicaudal D, regulates the branching of PVD sensory neuron dendrites. Development 2011, 138:507-518.
• [15]Dong X, Liu O, Howell A, Shen K: An extracellular adhesion molecule complex patterns dendritic branching and morphogenesis. Cell 2013, 155:296-307.
• [16]Maniar T, Kaplan M, Wang G, Shen K, Wei L, Shaw J, Koushika S, Bargmann C: UNC-33 (CRMP) and ankyrin organize microtubules and localize kinesin to polarize axon-dendrite sorting. Nat Neurosci 2012, 15:48-56.
• [17]Salzberg Y, Diaz-Balzac C, Ramirez-Suarez N, Attreed M, Tecle E, Desbois M, Kaprielian Z, Bullow H: Skin-Derived Cues Control Arborization of Sensory Dendrites in Caenorhabditis elegans. Cell 2013, 155:308-320.
• [18]Smith C, Watson J, Spencer W, O’Brien T, Cha B, Albeg A, Treinin M, MIller D: Time-lapse imaging and cell-specific expression profiling reveal dynamic branching and molecular determinants of a multi-dendritic nociceptor in C. elegans. Dev Biol 2010, 345:18-33.
• [19]Smith C, Watson J, Van-Hoven M, Colon-Ramos D, Miller D: Netrin (UNC-6) mediates dendritic self-avoidance. Nat Neurosci 2012, 15:731-737.
• [20]Smith C, O’Brien T, Chatzigeorgiou M, Spencer W, Feingold-Link E, Husson S, Hori S, Mitani S, Gottschalk A, Schafer W, MIller D: Sensory neuron fates are distinguished by a transcriptional switch that regulates dendrite branch stabilization. Neuron 2013, 79:266-280.
• [21]Mann A: Teams battle for neuron prize. Nature 2010, 467:143.
• [22]Nagarajan R: Intensity-based segmentation of microarray images. IEEE Trans Med Im 2003, 22:882-889.
• [23]Ramesh N, Yoo J, Sethi I: Thresholding based on histogram approximation. IEE Proc Vis Image Signal Process 1995, 142:271-279.
• [24]Lucieera A, Steina A, Fisherb P: Multivariate texture-based segmentation of remotely sensed imagery for extraction of objects and their uncertainty. Int J Remote Sensing 2005, 26:2917-2936.
• [25]Zwiggelaar R, Denton E: Texture, Based Segmentation. In Digital Mammography Volume 4046 of Lecture Notes in Computer Science. Edited by Astley S, Brady M, Rose C, Zwiggelaar R. Germany: Springer, Berlin Heidelberg; 2006:433-440.
• [26]Sharma N, Aggarwal L: Automated medical image segmentation techniques. J Med Phys 2010, 35:3-14.
• [27]Al-Kofahi K, Lasek S, Szarowski D, Pace C, Nagy G, Turner J, Roysam B: Rapid automated three-dimensional tracing of neurons from confocal image stacks. IEEE Trans Inform Technol Biomed 2002, 6:171-187.
• [28]Xiao L, Yuan X, Galbreath Z, Roysam B: Automatic and reliable extraction of dendrite backbone from optical microscopy images. In Life System Modeling and Intelligent Computing, Volume 6330 of Lecture Notes in Computer Science. Germany: Springer; 2010:100-112.
• [29]Peng H, Ruan Z, Long F, Simpson J, Myers E: V3D enables real-time 3D visualization and quantitative analysis of large-scale biological image data sets. Nat Biotechnol 2010, 24(4):348-353.
• [30]Gonzalez G, Türetken E, Fleuret F, Fua P: Delineating trees in noisy 2D images and 3D image-stacks. Computer Vision and Pattern Recognition (CVPR), IEEE Conference on, 2010, 2799-2806.
• [31]Turetken E, Benmansour F, Fua P: Automated reconstruction of tree structures using path classifiers and mixed integer programming. Computer Vision and Pattern Recognition (CVPR) 2012 IEEE Conference on, 2012.
• [32]Glowacki P, Pinhero M, Sznitman R, Turetken E, Lebrecht D, Holtmaat A, Kybic J, Fua P: Reconstructing evolving tree structures in time lapse sequences. IEEE Conference on Computer Vision and Pattern Recognition 2014. (in press)
• [33]Xiao H, Peng H: APP2: automatic tracing of 3D neuron morphology based on hierarchical pruning of gray-weighted image distance-trees. Bioinformatics 2013, 29(11):1448-1454.
• [34]Yuan X, Trachtenberg J, Potter S, Roysam B: MDL constrained 3-D grayscale skeletonization algorithm for automated extraction of dendrites and spines from fluorescence confocal images. Neuroinformatics 2009, 7(4):213-232.
• [35]Geusebroek J, Smeulders A, Weijer J: Fast anisotropic Gauss filtering. IEEE Trans Image Process 2003, 12:938-943.
• [36]Leung T, Malik J: Representing and recognizing the visual appearance of materials using three-dimensional textons. Int J Comput Vis 2001, 43:29-44.
• [37]Texture filter banks (Leung and Malik) http://www.robots.ox.ac.uk/~vgg/research/texclass/filters.html webcite
• [38]Schmid C: Constructing models for content-based image retrieval. Computer Vision and Pattern Recognition, 2001. CVPR 2001. Proceedings of the 2001 IEEE Computer Society Conference on, Volume 2, 2001, II-39–II–45.
• [39]Hastie T, Tibshirani R, Friedman J: The Elements of Statistical Learning. Springer Series in Statistics, Germany: Springer, second edition; 2009.
• [40]Metz C: Basic principles of ROC analysis. Semin Nuc 1978, 8:283-298.
• [41]Metz C: Receiver operating characteristic analysis: A tool for the quantitative evaluation of observer performance and imaging systems. J Am Coll Radiol 2006, 3:413-422.
• [42]Swets J: ROC analysis applied to the evaluation of medical imaging techniques. Invest R 1979, 14:109-121.
• [43]Kohavi R, Provost F: Glossary of terms. Mach Learn 1998, 30:271-274.
• [44]Duda R, Hart P: Use of the Hough transformation to detect lines and curves in pictures. Commun ACM 1972, 15:11-15.
• [45]Elfron B: Estimating the error rate of a prediction rule: improvement on cross-validation. J Am Stat Assoc 1983, 78:316-331.
• [46]Refaeilzadeh P, Tang L, Liu H: Encyclopedia of Database Systems. Germany: Springer; 2009 chap. Cross-Validation:. 532–538
• [47]Broser P, Erdogan S, Grinevich V, Osten P, Sakmann B, Wallace D: Automated axon length quantification for populations of labelled neurons. J Neurosci Methods 2008, 169:43-54.
• [48]Masseroli M, Bollea A, Forloni G: Quantitative morphology and shape classification of neurons by computerized image analysis. Comput Methods Programs Biomed 1993, 41:88-99.
• [49]Sznitman R, Gupta M, Hager G, Arratia P, Sznitman J: Multi-environment model estimation for motility analysis of Caenorhabditis elegans. PLOS One 2010, 5:e11631.
• [50]Gonzalez R, Woods R, Eddins S: Digital Image Processing using Matlab. USA: Gatesmark Publishing; 2009.
• [51]Image analysis & visualization system for bioimages & surface objects http://www.vaa3d.org/ webcite