Model-Resolution-based-Deconvolution-for-FD-Photoacoustic-Reconstruction

Model resolution based deconvolution (SALSA) improves frequency domain photo acoustic reconstruction. The full paper with detailed explanation and formulation can be found in https://doi.org/10.1121/10.0013829

Step 1 (Reconstruction_penalty_function_step1.m)

Parametres section

Input the parameters in Paramters section. Range of frequencies selected - (100kHz - 2.5Mhz) for FUNDUS and Derenzo Phantom and (5kHz - 150kHz) for Breast phantom. Set the required SNR and FOV. For split/ discontinuous frequency ranges, set fh_end, fh_start, wnh parameters.

Calculate W Matrix

Calculate Weight matrix (forward modelling) using function Calculate_W_matrix.m function. Regularization paramter for Derenzo phantom and FUNDUS is 100 and 2e11 for Breast phantom.

Calculate pressure matrix

Calculte P = WX to get pressure matrix (Complete forward model).

Inversion model using different penalty functions

Reconstruction is performed with a penalty function multiplied to regularization paramter.

Step 2 (Model_resolution_step2.m)

Set the same paramters as in step 1. All the deconvolutions were based on SALSA approach and were done for 100 iterations for each of the penalized reconstructions.