Traditional fast Fourier transform is used to extract the frequency component at the cost of losing the time domain, which is critical for metasurface biosensing. In this Letter, a more comprehensive algorithm, continuous wavelet transform (CWT), to process signals from THz time-domain spectroscopy is introduced. By comparing the metasurface-enhanced 2D time-frequency mappings (TFMs) of HaCaT and HSC3 cells, the two types of biological cells can be clearly differentiated, showing the great potential of CWT in the label-free recognition of biological cells. Also, the 2D TFMs serve as effective visualization indicators, successfully detecting the concentration of cancer cells characterized by being label free and low cost. In addition, the 2D TFMs of different metasurfaces under the same cell concentration reveal the correlation of TFMs and localized fields. Such a feature provides evidence of an interaction between biological cells and electromagnetic waves, implying the absorption of THz radiation by biological cells can be effectively controlled by properly designing split ring resonators (SRRs) of metasurfaces.