Electromagnetic signals propagating in dispersive and lossy media undergo additional attenuation when compared to their counterparts propagating in the free space. In this paper, we propose a threshold approach and Short-Time Fourier Transform (STFT) based inverse filters to compensate the attenuation in Time Reversal (TR) imaging algorithms. Although, the presence of dispersion and loss breaks the time reversal invariance of the wave equation in such media, the proposed algorithm allows improved TR focusing performance by minimizing the unwanted noise in the received signals without increasing the computational cost thanks to utilization of less number of inverse filters. The introduced algorithm can have applications in various disciplines such as medical imaging and subsurface detection.