Precise measurements of the composition of cosmic rays in the energy range of $\rm 10^{17} - 10^{18}\,eV$ could provide crucial insights into the long-standing questions about the origin and acceleration of these particles. Ground-based experiments typically rely on determining the position of the extensive air shower maximum ($X_\mathrm{max}$) to identify the type of cosmic ray particle. One effective method for determining $X_\mathrm{max}$ is by analyzing the radio emission produced by these air showers. This approach offers several advantages, including continuous operation and a higher duty cycle compared to fluorescence telescopes, which are limited by weather conditions and lunar phases. However, conventional radio-based methods often involve computationally intensive Monte Carlo simulations, or rely on pre-calculated parameterizations derived from simulations. In this contribution, we present a new method which is highly efficient and has the potential to reconstruct $X_\mathrm{max}$ with very minimal input from simulations. This method reconstructs the radio emission profile of air showers by backtracking the radio signals recorded by a ground-based antenna array, considering that the signal received by each antenna travels perpendicular to the radio wavefront. By analyzing simulated proton and iron showers in the $\rm 10^{17}-10^{18}\,eV$ range, this study reveals a strong correlation between the radio emission profiles in the 20–80 MHz frequency band, and the longitudinal profile of the air shower.