A commercial breast cancer detection system called T-Scan produces trans-admittance images (Neumann data) of breast tissues using a scan probe with a planar array of electrodes. The detection of a lesion in TAS is based on the difference between the trans-admittance data of the presence of anomaly and the absence of anomaly. However, the trans-admittance data of the absence of anomaly is not available in practice. To deal with this problem, we developed a multi-frequency trans-admittance scanner (TAS) for probing anomalies underneath a planar array of 320 electrodes from the frequency difference of the trans-admittance data. We present the mathematical analysis and numerical experiments showing the feasibility of multi-frequency lesion detection. Using the developed TAS system and saline phantoms, we conducted frequency-difference experiments.