Optimization with Partial Differential Equation Constraints for Bio-imaging Applications
Yiqian Wu,Yiwen Shi,P. Gill,Yingxiao Wang,Shaoying Lu
DOI: https://doi.org/10.51375/ijcvse.2021.1.3
2018-01-01
Abstract:Molecular transport and interaction are of fundamental importance in biology and medicine. The spatiotemporal diffusion map can reflect the regulation of molecular interactions and their intracellular functions. To construct subcellular diffusion maps based on bio-imaging data, we explore a general optimization framework with diffusion equation constraints (OPT-PDE). For the solution of the spatially piecewise constant and anisotropic diffusion tensors, we develop an efficient block solver based on apBank: The work of this author was supported by the National Science Foundation under contracts DMS-1318480 and DMS-1361421. Wu: The work of this author was supported in part by the National Science Foundation under contract DMS-1361421. Shi: The work of this author was supported in part by the National Science Foundation under contract DMS-1361421. Gill: The work of this author was supported by the National Science Foundation under contracts DMS-1318480 and DMS-1361421. Wang: The work of this author was supported by the National Science Foundation under contract DMS-1361421, and the National Institute of Health under contracts HL121365 and CA204. Lu: The work of this author was supported by the National Science Foundation under contract DMS-1361421, and the National Institute of Health under contracts HL121365 and CA204. Bank: Department of Mathematics, University of California, San Diego, La Jolla, California 92093-0112. Email:rbank@ucsd.edu Wu: Department of Bioengineering, Department of Mathematics, University of California, San Diego, La Jolla, California 92093-0435. Email:y5wu@ucsd.edu Shi: Department of Mathematics, University of California, San Diego, La Jolla, California 92093-0112. Email:yis018@ucsd.edu Gill: Department of Mathematics, University of California, San Diego, La Jolla, California 92093-0112. Email:pgill@ucsd.edu Wang: Department of Bioengineering, Department of Mathematics, University of California, San Diego, La Jolla, California 92093-0435. Email:yiw015@ucsd.edu Lu: Department of Bioengineering, Department of Mathematics, University of California, San Diego, La Jolla, California 92093-0435. Email:kalu@ucsd.edu plying Newton’s method to the first-order necessary condition for optimality. We characterize the wellposedness of the OPT-PDE model problem and the convergence properties of the solver. We also demonstrate the general utility of the solver in recovering spatially heterogeneous and anisotropic diffusion maps with computer-simulated bioimages. The results indicate that the solver can accurately recover piecewise-constant isotropic and anisotropic diffusion coefficients, while exhibiting efficient convergence and robustness. This work highlights the power of the OPT-PDE model and the solver in recovering the diffusion map from imaging data, and demonstrates that it has significant implications in bio-imaging analysis.