Introduction

In the linked pages, we present the code developed for analysis of binary images of polyp shapes and here discuss the rationale behind their development. The code for these programs was writtent in ALI (Analysis Language for Images) provided in the OPTIMAS software package, version 3.10 (BioScan Corporation, Redmond, WA).

Our goal in this endeavour was to characterize the dynamical behavior of hydrozoan polyps. Given sufficiently high temporal resolution of sampling, we could develop from these data a mathematical model to describe polyp behavior in which the parameters of the equations represent hypothesized biological features underlying their dynamics. Such a model is presented in Wagner et al. 1998.

The development of macros that could automate the capture and analysis of polyp images was clearly necessary from our need for high resolution temporal sampling and 3 features of polyp feeding behavior. First, digestion and transport of food through the gastrovascular system is a lengthy process. Regurgitation of ingested material by members of the colony may take up to ca. 24 hours to occur. Second, within the time between ingestion and regurgitation, fed polyps display characteristic behaviors that distinguish different phases of the digestion/transport process. Third, polyp oscillation behaviors that identify a phase of the feeding cycle occur on timescales of approx. every 2 minutes at 20ūC. Our exploratory work determined that adequate characterization of these 2 minute cycles of oscillation required sampling every 8 seconds (ca. 15-16 samples/cycle) to characterize the shape of the oscillation. Thus, in order to determine signature behaviors associated with each phase of feeding, the duration of each phase, and the characteristic behaviors associated with transitions between phases, continuous records of polyp dynamics were necessary, hence, the need for automation.

The earliest macros developed estimated only the dynamics of polyp length. Video records comparing isolated polyp behaviors to behaviors of polyps coupled to a colony demonstrated that measures of polyp length alone were inadequate to characterize polyp feeding behavior for the following reasons. Qualitative observations of length only in video records, as well as quantitative data on length dynamics, show similar patterns of behavior in isolated and coupled polyps. The differences between isolated and coupled polyps are masked by using only this metric. Video records show that when coupled polyps actively exchange gastrovascular fluid with the rest of the colony, polyps change shape most dramatically in width along the body column corresponding to the mid-gastric region. These oscillations in width effect large amplitude oscillations in polyp volume, in some cases exceeding 100% change in polyp volume with each cycle (e.g., a 6-8 nl polyp that balloons to approx. 18 nl following import of fluid). Such large amplitude volume changes do not occur in isolated polyps.

We developed macros that measured both length and 11 widths evenly spaced across the long axis of the polyp. From these measures, we estimated volume of the polyp for each image sampled by assuming rotational symmetry of polyps and applying the extended Simpson's rule (Press et al. 1992).

The macros for analyzing polyp shape consist of two macros, one nested inside the other. Both are modularly constructed by defining sub-routines within the code that are later called upon to perform their operation. The first macro, runvol6.mac, is the control program that sets the configuration, data file export parameters, test measures, and the control of cyclic processing of stored images. The second macro, scan6.mac, is nested within runvol6.mac and contains the code for routines that actually make the measurements. Both macros are fully annotated to explain the procedure carried out by the code that follows the annotation. Variants of these macros that differ slightly in how noise is filtered from stored images, or how many widths along the body column are sampled, etc., as well as the code to automate the sampling of images from videotape are also available by contacting me.

References


Press, W.H., S.A. Teukolsky, W.T. Vetterling, and B.P. Flannery. 1992. Numerical Recipes in C. Cambridge University Press, New York.
Wagner, A., S. R. Dudgeon, J. R. Vaisnys, and L. W. Buss. 1998. Non-linear oscillations in polyps of the colonial hydroid, Podocoryne carnea. Naturwissenschaften 85: 117-120.

steve.dudgeon@csun.edu
Gastrovascular Dynamics
Dudgeon Lab Home
CSUN-Biology Home
© 1999 Steve Dudgeon
Page last updated March 19, 1999