One key component to accurate photoacoustic image reconstruction with flexible transducers is knowledge of element positions as the flexible array shape varies. Our previously proposed metric calculates the maximum lag-one spatial coherence (mLOC) of a photoacoustic target, with demonstrated advantages to estimate the curvature of a flexible array. However, the effectiveness of mLOC in relation to target locations remained unclear. Therefore, we investigated mLOC performance across different source locations by first designing a radial grid consisting of 20 simulated photoacoustic sources at varying azimuthal angles and depths. The accuracy of the mLOC-estimated array radius was 97.59%-100% when compared with the ground truth curvature. Signals from deeper sources produced greater coherence when beamforming images with the correctly estimated array radius. In addition, we introduced a normalized full-width at half-prominence (nFWHP) metric to assess the flatness of the mLOC measurements as a function of estimated array radius, demonstrating greater mLOC effectiveness with estimating flexible array radii with decreasing distance between the source and the transducer center.
