Title: Virtual Fruit tissue Generation Based on Cell Growth Modeling
Other Titles: Virtuele weefselgeneratie van fruit op basis van celgroeimodellen
Authors: Abera, Metadel
Issue Date: 14-Jan-2014
Abstract: The fact that production of pome fruits is both season and location dependent calls for preservation methods to maintain the quality of the produce after harvest. The principal method to preserve pome fruit is postharvest storage in cool rooms under Controlled Atmosphere (CA) or Ultra Low Oxygen (ULO) conditions. These methods are based on empirical trials to determine optimum storage conditions for a specific cultivar in terms of temperature, oxygen and carbon dioxide concentration and relative humidity. Improved understanding of the underlying fruit physiology in relation to the gas and water exchange processes during storage will assist in improving postharvest quality and reducing the occurrence of storage disorders. Both experimental and modeling approaches have been followed to investigate the relationship between the gas concentration, gas diffusion, respiration and physiological disorders in apple and pear. Experimental approaches are, however, costly, tiresome and time consuming. Moreover, it is difficult to investigate the time course of the physiological disorders experimentally because of the lack of non-destructive techniques to measure the internal gas concentration in the fruit. Alternatively, mathematical models can be used to study gas and water exchange. They are usually based on the continuum hypothesis where the fruit is considered as a material with transport properties that are independent of the spatial scale. However, unlike the traditional engineering materials fruit tissue has a complex fine structure. The cellular architecture is believed to determine to a large extent the biophysical processes in the fruit. The continuum hypothesis does not hold in this case and a multiscale approach is required in which the model parameters of the model that operates at the macroscale – the scale of interest – are obtained from simulations with a microscale model that incorporates the actual microstructure of the fruit. For the latter, microscale geometric models of the fruit are required. Pome fruit tissue microscale geometry generators exist today but are based on digitized 2D or 3D images of the cellular architecture. Therefore, although these algorithms generate representative geometries of the tissues, they require experimental input in terms of microscopic images. These approaches need complex image acquisition procedures and expensive infrastructures such as synchrotron radiation sources. Also, they do not allow to parameterize the microscale geometry to, for example, investigate the effect of cell size or shape on the transport properties in a systematic way. The main objective of this dissertation was to develop virtual microscale fruit tissue generators (algorithms) that generate statistically and spatially equivalent virtual tissue microstructures resolving the cell symplast, cell wall and intercellular air spaces in both 2D and 3D and interface them to finite element and/or finite volume codes. To achieve this, we have developed virtual tissue generators that are based on cell growth modeling bytaking into account cell biomechanics. The generators are initiated from a random Voronoi tessellation and growth biomechanics is applied to the tessellation which results in a virtual tissue that has equivalent geometrical properties as that of real tissues obtained from microscopic or synchrotron microtomography images. In a further extension we have also developed a cell division algorithm which is based on cell biomechanics and that is capable of mimicking both symmetric cell division and asymmetric cell division with different degree of anisotropic growth. The cell division algorithm can be used instead of the Voronoi tessellation as an input for the expansive growth models. Initial tessellations obtained from the cell division algorithm will have more realistic representation of the cells than the Voronoi tessellations. The geometric models can be used to carry out in silico simulations to determine transport properties to be used in multiscale framework of gas and moisture exchange studies in pome fruits. This approach helps to include more geometrical details and fewer assumptions than the classical continuum modeling approach, while requiring less computer time compared to solving governing model equations at the resolution of the microscale.
Table of Contents: Acknowledgements i
Abstract iii
Beknopte samenvatting vii
Table of contents xi
Abbreviations and symbols xv
1 General introduction 1
1.1 Introduction 1
1.2 Modeling approaches of transport phenomena in fruit tissues 5
1.2.1 The need for modeling transport phenomena in fruit tissues 5
1.2.2 Continuum modeling of transport phenomena in fruit tissues 8
1.2.3 Importance of microstructure in modeling transport phenomena in fruit tissues 10
1.3 Objectives 11
1.4 Thesis outline 12
Bibliography 13
2 Literature review 21
2.1 Introduction 21
2.2 Relevance of geometry of fruit tissue at different scales 22
2.3 Fruit tissue structure 26
2.3.1 Plant tissue types 26
2.3.2 Meristematic versus permanent tissues 26
2.3.3 Pome fruit tissue types 28
2.4 Microscale fruit tissue models based on imaging 32
2.5 Microscale fruit tissue modeling using tesselations 34
2.6 Microscale fruit tissue geometries based on tissue growth models 37
2.7 Conclusions 43
Bibliography 44
3 2D virtual fruit tissue generation based on cell growth modeling 53
3.1 Introduction 53
3.2 Material and methods 54
3.2.1 Sample preparation, image acquisition and processing 54
3.2.2 The growth model 55 Governing equations 55 Model parameters 58 Algorithm 59
3.2.3 Sensitivity analysis 61
3.2.4 Statistical analysis 62
3.3 Results and discussion 63
3.3.1 Parenchyma tissue 63
3.3.2 Complex tissue structures 69
3.3.3 Applications 71
3.4 Conclusions 72
Bibliography 73
4 3D virtual fruit tissue generation based on cell growth modeling 77
4.1 Introduction 77
4.2 Materials and methods 79
4.2.1 The growth model 79 Governing equations 79 Model parameters 83 Algorithm 83 Pore generation 86
4.2.2 Sample preparation, image acquisition and processing 86
4.2.3 Sensitivity analysis 87
4.2.4 Statistical analysis 87
4.3 Results 88
4.3.1 Cell size and shape 89
4.3.2 Pore generation 90 Schizogenous origin pore generation 90 Lysigenous origin pore generation 91
4.3.3 Calibration of the model using image data 92
4.3.4 Complex tissue structures 96 The skin 96 Stone cells 98
4.4 Discussion 99
4.5 Conclusions 103
Bibliography 103
5 2D cell division algorithm based on ellipse fitting 109
5.1 Introduction 109
5.2 Materials and methods 112
5.2.1 Cell growth algorithm 112
5.2.2 Cell division algorithm 114
5.2.3 Experimental data 115
5.2.4 Geometric and topological properties 116 Topology 116 Cell shape 116 Cell size 118 Statistical comparison 118
5.3 Results 118
5.3.1 Topology distribution 120
5.3.2 Cell size distribution 122
5.3.3 Cell shape distribution 123 Aspect ratio distribution 124 Internal angle distribution 126
5.3.4 Comparison of real and virtual topological and geometrical properties 128
5.4 Discussions 131
5.5 Conclusions 135
Bibliography 136
6 3D cell division algorithm based on ellipsoid fitting 141
6.1 Introduction 141
6.2 Materials and methods 144
6.2.1 Cell growth algorithm 144
6.2.2 Cell division algorithm 146
6.2.3 Geometric and topological properties 148
6.2.4 Statistical comparison 149
6.3 Results 149
6.3.1 Topology distribution 151
6.3.2 Cell size distribution 153
6.3.3 Cell shape distribution 154
6.4 Discussion 156
6.5 Conclusions 159
Bibliography 160
7 General conclusion and future directions 165
7.1 General conclusions 165
7.2 Future directions 168
Bibliography 169
List of publications 173
ISBN: 978-90-8826-344-6
Publication status: published
KU Leuven publication type: TH
Appears in Collections:Division of Mechatronics, Biostatistics and Sensors (MeBioS)

Files in This Item:
File Status SizeFormat
DISSERTATION _Metadel ABERA.pdf Published 13018KbAdobe PDFView/Open Request a copy

These files are only available to some KU Leuven Association staff members


All items in Lirias are protected by copyright, with all rights reserved.