Hybrid Multiscale Modeling of Cancer Cell Behavior

Zangooei, Mohammad Hossein | 2017

455 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 50840 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Habibi, Jafar
  7. Abstract:
  8. Cancer is a class of diseases characterized by out-of-control cell growth. Cancer is among the leading causes of death worldwide.Cancer modeling is increasingly being recognized as a powerful tool to refine hypotheses, focus experiments, and enable predictions that are more accurate.We investigate a three-dimensional multiscale model of vascular tumour growth, which couples blood flow, angiogenesis, vascular remodelling, nutrient/growth factor transport, movement of, and interactions between, normal and tumour cells. We constructed a hybrid multi- scale agent-based model that combines continuous and discrete methods.Each cell is represented as an agent. The agents have rules that they must follow in the course of a simulation, both for their independent behavior and for inter- actions between other agents. We make an evolutionary game theory component in order to model the signal propagation through gap junctions.Most current cancer modeling approaches lead to rough estimates of outcomes and their results cannot conform to individual patients because number of cancerous/healthy cells is based on random initialization. We design a classification approach for histopathological images. In light of these problems, we have constructed a cell-graph model based on the spatial relationship between the cells to quantify the properties of cancer cells. After constructing the cell-graphs and extracting a rich set of features (image-based and graph-based), a sophisticated learning technique is applied to determine the type of cells. In this article, we attack the classification problem using the regression approach equipped with kernel fusion technique. Since this method aggregates kernels based on NSGA-II, we called it Hybrid SVR-NSGA-II.
    Computational simulations of the model were performed in order to analyze its performance. The most striking feature of our results is that each cell can select its phenotype at each time step according to its condition. We provide evidence that the prediction of cell phenotypes is reliable.Our proposed model, which we term a hybrid multiscale modeling of cancer cell behavior, has the potential to combine the best features of both continuum and discrete models. The in silico results indicate that the 3D model can represent key features of cancer growth, angiogenesis, and its related micro-environment and show that the findings are in good agreement with biological tumor behavior. To the best of our knowledge, this paper is the first hybrid vascular multiscale modeling of cancer cell behavior that has the capability to predict cell phenotypes individually by a self-generated dataset
  9. Keywords:
  10. Multiscale Modeling ; Graphs ; Machine Learning ; Cancer ; Support Vector Machine (SVM) ; Agent Based Model

 Digital Object List