Position Based Single Particle Tracking (P-SPT) for High-Resolution Deformation and Motion Capture

Abstract

Our recently developed Position based Single Particle Tracking (P-SPT) algorithm utilizes a novel approach to track particles at high seeding density in 2D and 3D at very fast computational speeds. Traditional kinematic measurement technique like single particle tracking (SPT) struggle at accurately reconstructing large, nonlinear material deformation fields, and high spatial frequency content as they are computationally very expensive for high particle seeding densities, especially in 3D. Despite recent advances, approaches like image correlation are also computationally expensive and are limited in the recovery of high spatial frequency information due to their intrinsic correlation nature. P-SPT tracks particles by creating unique feature identifiers for each particle using the relative geometric position of neighboring particles. This unique feature identifier, or descriptor, is used for particle linking between two successive time frames. The nature of feature descriptor along with a deformation warping function enables P-SPT to capture large, nonlinear material deformation fields. Validation of P-SPT performed on synthetically generated 3D volumetric images show that P-SPT is computationally efficient equipping it to track particles at high seeding densities far exceeding the typical bead densities used in traditional SPT methodologies thus enabling it to reconstruct high spatial frequency information. Since P-SPT only requires the particle position for tracking, it can be employed for tracking motion fields in 2D images, 3D volumetric images or multi-camera imaging system. This flexibility of P-SPT along with its computational efficiency means that it can be executed reliably on a simple personal computers in variety of biological applications that require reconstruction of high frequency and high spatial motion content, like quantifying cell deformations for 2D and 3D traction force microscopy, subcellular dynamics or particle dynamics in granular and cellular media.