# Installing CULA (txt|pdf) -for using mex files
Matlab Executables:
JTJ Routine (Update Equation 3): calc_update_gpu_JTJ.mexa64
JJT Routine (Update Equation 4): calc_update_gpu_JJT.mexa64
# Installing ViennaCL (txt|pdf) -for using mex files
Matlab Executable:
Sparse Finite Element Solver: viennacl_bicgstab_precon.mexa64
# Creation of MEX files using CULA functions (txt|pdf)
Example: Mex wrapper for culasv code: culasv.c
Mex wrapper for ViennaCL code: viennacl_bicgstab_precon.cpp
# Matlab Code for Image reconstruction (requires NIRFAST)
On GPU:
Jacobian computation: get_field_GPU_inv.m
Reconstruction computation: reconstruction_stnd_GPU_inv.m
Note: These instructions are for linux operating system only.
This Matlab code is used as part of the work presented in:
J. Prakash, V. Desai, S. Srinivasan, and P. K. Yalavarthy, "Multi-core computers have high scalability than graphics processing units for diffuse optical tomographic image reconstruction," SPIE/OSA European Conference on Biomedical Optics (ECBO-2013), May 12-16, 2013, Munich, Germany
Created on: October 13, 2013.