Electrical impedance tomography (EIT) is a safe and promising imaging technique used in the medical field. Its non-invasive nature offers significant potential for various diagnostic applications. However, the sensitivity of EIT renders it susceptible to environmental influences, which directly affect the quality of image reconstructions. Moreover, factors such as frequency, current intensity, and choice of reconstruction algorithm are pivotal in determining the efficacy of image reconstruction, underscoring the importance of selecting an appropriate algorithm tailored to a specific survey frequency. This study conducted experiments using a test phantom model equipped with aluminum metal conductors, exploring frequencies ranging from 1 to 220 kHz. Reconstruction algorithms have been employed, including Jacobian, Gauss-Newton, Tikhonov, NOSER, and filter back projection (FBP). Quantitative metrics such as position error (PE), resolution (RES), contrast-to-noise ratio (CNR), and Sharpness were utilized to comprehensively evaluate the performance of the algorithms and the influence of frequency on image reconstruction outcomes. The analysis revealed a significant correlation between the frequency and effectiveness of image reconstruction algorithms. Notably, the algorithms exhibited high performance within the 10–100 kHz frequency range, surpassing a normalized threshold value of 0.9. Although the NOSER regularization method exhibits stable and fine performance within the surveyed frequency range, it still causes artifacts in the EIT images. Although this study sheds light on the challenges and nuances associated with image reconstruction in EIT, particularly concerning convergence and the tradeoff between PE and CNR, it underscores the critical importance of algorithm selection tailored to specific imaging requirements and frequencies. Overcoming these challenges and refining reconstruction methodologies will be pivotal for enhancing the clinical utility and efficacy of EIT in medical imaging applications.