Conventional rock-physics models often neglect the coupled effects of kerogen morphology and pore geometry on elastic-wave propagation when estimating shear-wave velocity in organic-rich rocks. To address this limitation, this study proposes a shear-wave velocity prediction method that integrates the Kuster–Toksöz model with a particle swarm optimization (PSO) algorithm. In this framework, organic-rich rock is represented as a three-component composite medium consisting of a mineral matrix, kerogen inclusions, and fluid-filled pores. Both kerogen and pores are characterized using two idealized inclusion geometries, namely spherical and penny-shaped forms. The corresponding volume fractions are defined as morphological parameters to quantify their influence on the elastic properties of the rock. An objective function is established by minimizing the misfit between predicted and measured P-wave velocities, and the resulting nonlinear multi-parameter inversion problem is solved using the PSO algorithm to achieve global optimization. During the inversion process, the volume fractions of spherical and penny-shaped kerogen inclusions, as well as those of pores, are simultaneously estimated. The optimized morphological parameters are then used to predict shear-wave velocity. The proposed method is validated using both laboratory core measurements and well-log data, and its performance is compared with that of two existing shear-wave velocity prediction methods based on single-parameter calibration. The results demonstrate that the proposed framework significantly improves prediction accuracy over the reference methods, highlighting its potential for reliable shear-wave velocity estimation in organic-rich rocks.