Rapid Personalized Computational Modeling of the Wrist - Data

Project Description

This work was performed at Mayo Clinic in Rochester, MN, USA.

It is part of a bigger project to use four-dimensional computed tomography data to develop and calibrate computational models of the human hand and wrist for recreating joint dynamics, with the ultimate goal being to improve the diagnosis and treatment of soft-tissue injuries of the hand and wrist. The work is funded by several NIH grants, as well as internal funding from Mayo Clinic. The project includes a data collected from a previous study on in vivo healthy controls collected with previous IRB approval (NCT03193996) using a combination of static computed tomography (3DCT) and dynamic computed tomography (4DCT = 3DCT + time). In this case, a subset of three asymptomatic individuals from that dataset were used. A novel computational workflow was used that combined non-linear morphing algorithms based on generalized regression neural networks (GRNNs) with algorithmic techniques to automatically create computational finite element models of each individuals native geometries. This workflow was applied to the three asymptomatic participants to develop models with personalized wrist bone geometries and material properties. Each model was then used with an optimization routine as well as a Monte Carlo (MC) analysis to demonstrate the utility of these models towards developing Digital Twins and for in silico clinical trials.

In total the work has provided the raw geometries and kinematics for the three participants, the calibrated computational model for Participant 2, and the working MATLAB code to develop new personalized models from the original bony geometries. This work is shared with IRB approval (25-012254).

This repository contains all of the raw data including the geometries and the kinematics for the three participants. The full set of deliverables are including in the following links:

Data: www.doi.org/10.5281/zenodo.18166549

Model: www.doi.org/10.5281/zenodo.18166611

Software/Code: www.doi.org/10.5281/zenodo.18166629

Licensing

This work is provided under a Creative Commons Attribution 4.0 International. You are free to use the work for your own purposes; however, we ask that you cite the work that made it possible:

Thor E. Andreassen, Taylor P. Trentadue, Andrew R. Thoreson, Kai-Nan An, Sanjeev Kakar, Kristin D. Zhao, "Rapid Development of Efficient Participant-Specific Computational Models of the Wrist", arXiv, 25 May 2025, DOI: 10.48550/arXiv.2505.19282

and

Thor E. Andreassen, Taylor P. Trentadue, Andrew R. Thoreson, Kai-Nan An, Sanjeev Kakar, Kristin D. Zhao, "Rapid Personalized Computational Modeling of the Wrist - Data", Zenodo, 14 January 2026, DOI: 10.5281/zenodo.18166549

File Structure

All files are organized within folders and name accordingly. For descriptions of the files and folder, please download the README.txt file to see the details.