The zebrafish offers unique advantages as a vertebrate system to analyze cell and tissue structure and development in vivo. External fertilization and transparent larvae allow imaging of cells in live, developing animals. New genetic techniques can knock out most zebrafish genes, making it critical to develop high- throughput techniques to analyze mutant phenotypes (the """"""""phenome project""""""""). For all these experiments, confocal microscopy is an essential tool, making it equally critical to develop software to analyze 3D and 4D (3D over time) confocal data and quickly derive biologically-relevant conclusions, both qualitative and quantitative. Image analysis requires five steps: preprocessing, registration, visualization, segmentation, and quantitation. We have already developed a downloadable confocal analysis application, FluoRender, and optimized it for visualization, providing significant advantages over existing commercial applications. The present proposal would make FluoRender a full-featured package by adding the other four steps. Our goal is to enable the biologist to easily analyze 3D and 4D confocal data, identifying and measuring features of interest, and allowing rapid repetitive analysis of multiple samples. For features like segmentation that benefit from automation, we provide user validation and editing, emphasizing high accuracy rather than pure speed. We will focus on three specific problems: 1) 3D image mosaicking, 4D drift removal;2) 3D segmentation of confocal data;and 3) 4D tracking in confocal data. Mosaicking allows scanning of specimens larger than a microscope's field of view. Timelapse experiments benefit from 4D drift removal, since the growing embryo can change shape or the microscope's focus can drift. We will develop segmentation methods for objects from three classes: nuclei/cells, axons/dendrites, and tissues. We will develop semi-automatic segmentation methods for zebrafish labeled with multiple fluorophores, or spectrally (Brainbow). We will develop methods so that once cells or tissues are initially segmented, they can be tracked over time, then visualized co-registered with the raw data, providing context for interpreting their motion. The improved FluoRender software will provide a freely-distributed, portable suite that enables rapid analysis of 3D or 4D confocal datasets. This will aid in analyzing cell movements, neuronal circuitry, and tissue development in wildtype and mutant embryos. Given the growing relevance of zebrafish as a human disease model, the proposed analysis software can be expected to benefit our understanding of many different developmental, neurobiological, and metabolic diseases.

Public Health Relevance

The zebrafish is becoming widely used as a model of developmental, neurological, and physiological diseases in humans. An important strength of this model system is the ability to use confocal microscopy and related techniques to take three-dimensional (3D) and four-dimensional (3D over time, or 4D) images, showing the positions and structures of cells and organs in the 3D context of living zebrafish embryos and larvae. This proposal would develop new 3D/4D image analysis software that would help to take advantage of these powerful imaging techniques to understand in detail the defects that can occur in nervous system or organ structure and development, with the long-term goal of understanding related defects in human disease.

National Institute of Health (NIH)
National Institute of General Medical Sciences (NIGMS)
Research Project (R01)
Project #
Application #
Study Section
Special Emphasis Panel (ZRG1-CB-Z (56))
Program Officer
Deatherage, James F
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Utah
Organized Research Units
Salt Lake City
United States
Zip Code
Jezzini, Sami H; Merced, Amelia; Blagburn, Jonathan M (2018) Shaking-B misexpression increases the formation of gap junctions but not chemical synapses between auditory sensory neurons and the giant fiber of Drosophila melanogaster. PLoS One 13:e0198710
Mesadi, Fitsum; Erdil, Ertunc; Cetin, Mujdat et al. (2018) Image Segmentation Using Disjunctive Normal Bayesian Shape and Appearance Models. IEEE Trans Med Imaging 37:293-305
Mesadi, Fitsum; Cetin, Mujdat; Tasdizen, Tolga (2016) DISJUNCTIVE NORMAL LEVEL SET: AN EFFICIENT PARAMETRIC IMPLICIT METHOD. Proc Int Conf Image Proc 2016:4299-4303
P├ęzier, Adeline P; Jezzini, Sami H; Bacon, Jonathan P et al. (2016) Shaking B Mediates Synaptic Coupling between Auditory Sensory Neurons and the Giant Fiber of Drosophila melanogaster. PLoS One 11:e0152211
Takahashi, Akiko; Islam, M Sadiqul; Abe, Hideki et al. (2016) Morphological analysis of the early development of telencephalic and diencephalic gonadotropin-releasing hormone neuronal systems in enhanced green fluorescent protein-expressing transgenic medaka lines. J Comp Neurol 524:896-913
Holman, Holly A; Tran, Vy M; Kalita, Mausam et al. (2016) BODIPY-Conjugated Xyloside Primes Fluorescent Glycosaminoglycans in the Inner Ear of Opsanus tau. J Assoc Res Otolaryngol 17:525-540
Zhou, Liang; Hansen, Charles D (2016) A Survey of Colormaps in Visualization. IEEE Trans Vis Comput Graph 22:2051-69
Ando, Koji; Fukuhara, Shigetomo; Izumi, Nanae et al. (2016) Clarification of mural cell coverage of vascular endothelial cells by live imaging of zebrafish. Development 143:1328-39
Figueroa, Francisco; Singer, Susan S; LeClair, Elizabeth E (2015) Making maxillary barbels with a proximal-distal gradient of Wnt signals in matrix-bound mesenchymal cells. Evol Dev 17:367-79
Mesadi, Fitsum; Cetin, Mujdat; Tasdizen, Tolga (2015) Disjunctive Normal Shape and Appearance Priors with Applications to Image Segmentation. Med Image Comput Comput Assist Interv 9351:703-710

Showing the most recent 10 out of 23 publications