Home J!Research Publications Novo

Mining Data from Hemodynamic Simulations for Human Aorta

Research Area: Uncategorized Year: 2011
Type of Publication: Article Keywords: knearest neighbor, multilayer perceptron neural network, backpropagation algorithm, CFD simulatio
  • Tosovic, M.
  • Radovic, Milos
  • Petrovic, Dejan
  • Milasinovic, Danko
  • Devedzic, Goran
  • Rankovic, Vesna
  • Filipovic, Nenad
Journal: IPSI BgD journals Volume: 7
Number: 2
To analyze the influence of geometric parameters and flow on maximal wall shear stress (MWSS), maximum pressure and total force in the human aorta, the computer simulations were run to generate the data pertaining to this phenomenon. These data were used for training and testing two data mining models: multilayer perceptron neural network trained with backpropagation algorithm and k-nearest neighbor algorithm. The results showed that both models have strong ability in predicting target variables