Computational & Technology Resources
an online resource for computational,
engineering & technology publications
engineering & technology publications
Civil-Comp Conferences
ISSN 2753-3239
ISSN 2753-3239
CCC: 11
PROCEEDINGS OF THE SEVENTH INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE, SOFT COMPUTING, MACHINE LEARNING AND OPTIMIZATION IN ENGINEERING
PROCEEDINGS OF THE SEVENTH INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE, SOFT COMPUTING, MACHINE LEARNING AND OPTIMIZATION IN ENGINEERING
Edited by: P. Iványi, J. Kruis and B.H.V. Topping
Paper 1.2
Conceptualizing an AI-based Effective Stiffness Analysis of Human Trabecular Bone
J. Gebert1,2, F. Pelzer1,2 and M.M. Resch1,2
1, High-Performance Computing Center Stuttgart, Germany
2Institute for High-Performance Computing, University of Stutttgart, Germany
Full Bibliographic Reference for this paper
J. Gebert, F. Pelzer, M.M. Resch, "Conceptualizing an AI-based Effective Stiffness Analysis of Human Trabecular Bone", in P. Iványi, J. Kruis, B.H.V. Topping, (Editors), "Proceedings of the Seventh International Conference on
Artificial Intelligence, Soft Computing, Machine Learning and Optimization in Engineering", Civil-Comp Press, Edinburgh, UK,
Online volume: CCC 11, Paper 1.2, 2025, doi:10.4203/ccc.11.1.2
Keywords: direct mechanics, finite elements, human bone, high-performance computing, artificial intelligence, effective stiffness, stiffness.
Abstract
Linear elastic material characterization of human trabecular bone is of interest, for example, to simulate bone-implant systems. The direct mechanics method for computing stiffness tensors of trabeculae based on directly discretized computed tomography scans is proven and available. However, characterizing the effective stiffness based on microfocus CT scans is computationally expensive. We suggest a new, AI-based method for characterizing volume elements used in direct mechanics approaches. The goals are to reduce the required computational effort and cost while keeping a comparable accuracy of the effective stiffness parameters that we can compare to the existing analytical ground truth. Our approach assumes that binary segmented grayscale CT images are 3-dimensional patterns that answer deformations with particular forces and exhibit a specific effective stiffness. The paper describes the assumptions, training data, and software stack for AI training and inference. We conclude with expected results, methodological limits, and an outlook.
download the full-text of this paper (PDF, 6031 Kb)
go to the previous paper
go to the next paper
return to the table of contents
return to the volume description