Morphological kinetics and distribution in somatic embryo cultures

Abstract

The environmental effects on developing somatic embryos should be characterized not only by the growth based on biomass, but also by the morphological properties and size. We have previously developed a discrete classifier to separate developing embryos into distinct morphological classes. In this study, a continuous descriptor using the distributions of magnitude of features representing morphological characteristics and size information was used to describe the developing embryo populations. The identity of the population was examined by comparing either the distributions of all features or key features. The method was applied to characterize the kinetics of carrot embryo populations cultivated in the presence and absence of triiodobenzoic acid(TIBA), an inhibitor of auxin polar transport. Optimal sample size for morphological characterization was determined by the invariance of feature distributions with further increase in sample size. The overall growth and substrate consumption kinetics were only slightly affected by the presence of TIBA. However, the distribution of morphological features was significantly affected. The features showing the highest statistical significance were related to those corresponding to the roughness. The continuous descriptor for characterizing developing embryo population is potentially useful for quality control in large-scale operations.

Effects from shear stress on morphology and growth of early stages of Norway spruce somatic embryos

“…Furthermore, spectral analysis through discrete and fast Fourier transformations has been applied to acquire quantitative information on the presence of different embryo shapes in culture (Cazzulino et al, 1987;Chi et al, 1994Chi et al, , 1996Vits et al, 1994). The features converted from the segments of an embryo contour by discretizing the entire contour with an equal angle increment, resulted in a non-uniform contour sampling and could not consistently define the cotyledonary embryos .…”

Classification of Plant Somatic Embryos by Using Neural Network Classifiers

“…These features were grouped into four categories according to their geometric meaning: (a) maturation index, polarization (Fourier feature -3) and elongation (Fourier feature +1); (b) size features, perimeter, area, and biovolume; (c) asymmetry, Fourier features -2 and +2; and (d) complexity, the average value of the magnitude of medium-range Fourier features (-8 to -4 and +4 to +8). The geometric significance of each feature has been described previously (9). The two pattern-recognition systems used similar features in their classification as can be seen from Figures 2 and 3.…”

“…Geometrical Interpretation of the Fourier Features. Each Fourier coefficient obtained as described above has its own geometrical meaning that characterizes certain morphological components of the somatic embryos (9). However, not all 32 Fourier features are important in characterizing the morphology of the embryos.…”

“…In the last decade, much progress has been made in the development of image analysis and pattern-recognition systems to classify somatic embryos (4)(5)(6)(7)(8). We have been employing such a system to classify both normal and abnormal carrot somatic embryos into various classes (9,10). A video camera is used to capture embryo images first.…”

Characterizing patterns in plant somatic embryo cultures: Its morphology and development

“…The resulting correlation with intrinsic developmental phenomena at the metabolic and mechanical levels is indirect. Therefore, the experimental verification of proposed mechanistic hypothesis (Vits et al, 1992) employing image processing techniques must rely on statistical inference methods (Chi et al, 1994).…”