This paper presents a near-field beam tracking framework for dynamic metasurface antennas (DMAs) in high-frequency wireless communication systems. The framework initiates beam sweeping only when the base station (BS) estimates that its beamforming gain drops below a threshold from its theoretically optimum value. The paper introduces novel analytical expressions for the correlation function between beam focusing vectors, beamforming gain with respect to user coordinate mismatch, the direction of user movement yielding the fastest beamforming gain deterioration, and the minimum user displacement for a certain performance loss. A non-uniform coordinate grid is designed to effectively sample the user area of interest at each position estimation slot. The framework leverages the effective beam coherence time metric, which indicates the minimum time needed for the user to experience a specific loss relative to the theoretically optimal beam focusing gain. The paper validates the proposed framework through extensive simulations, demonstrating its superiority over benchmarks. The results show that the proposed near-field beam tracking framework is effective in maintaining high beamforming gain in dynamic environments. The framework is designed for a generic scenario where the mobile user lies in a plane vertical to the BS plane with a constant height difference. The paper also provides analytical expressions for the beamforming gain under UE coordinate mismatches, including range and azimuth angle mismatches. The results show that the proposed framework can achieve high beamforming gain even for large distances, extending beyond the Rayleigh distance. The framework is designed to dynamically adjust the sampling resolution based on the user's estimated position and movement. The paper also introduces a hybrid analog and digital beam scanning approach, which allows for efficient sampling of the user area of interest. The framework is validated through simulations, demonstrating its effectiveness in maintaining high beamforming gain in dynamic environments. The paper concludes that the proposed near-field beam tracking framework is a promising solution for high-frequency wireless communication systems.This paper presents a near-field beam tracking framework for dynamic metasurface antennas (DMAs) in high-frequency wireless communication systems. The framework initiates beam sweeping only when the base station (BS) estimates that its beamforming gain drops below a threshold from its theoretically optimum value. The paper introduces novel analytical expressions for the correlation function between beam focusing vectors, beamforming gain with respect to user coordinate mismatch, the direction of user movement yielding the fastest beamforming gain deterioration, and the minimum user displacement for a certain performance loss. A non-uniform coordinate grid is designed to effectively sample the user area of interest at each position estimation slot. The framework leverages the effective beam coherence time metric, which indicates the minimum time needed for the user to experience a specific loss relative to the theoretically optimal beam focusing gain. The paper validates the proposed framework through extensive simulations, demonstrating its superiority over benchmarks. The results show that the proposed near-field beam tracking framework is effective in maintaining high beamforming gain in dynamic environments. The framework is designed for a generic scenario where the mobile user lies in a plane vertical to the BS plane with a constant height difference. The paper also provides analytical expressions for the beamforming gain under UE coordinate mismatches, including range and azimuth angle mismatches. The results show that the proposed framework can achieve high beamforming gain even for large distances, extending beyond the Rayleigh distance. The framework is designed to dynamically adjust the sampling resolution based on the user's estimated position and movement. The paper also introduces a hybrid analog and digital beam scanning approach, which allows for efficient sampling of the user area of interest. The framework is validated through simulations, demonstrating its effectiveness in maintaining high beamforming gain in dynamic environments. The paper concludes that the proposed near-field beam tracking framework is a promising solution for high-frequency wireless communication systems.