A new model is provided for understanding and exploiting impedance or admittance values measured by implantable medical devices, such as pacemakers or cardiac resynchronization devices (CRTs.) The device measures impedance along vectors extending through tissues of the patient between various pairs of electrodes. The device then converts the vector-based impedance measurements into near-field individual electrode-based impedance values. This is accomplished, in at least some examples, by converting the vector-based impedance measurements into a set of linear equations to be solved while ignoring far-field contributions to the impedance measurements. The device solves the linear equations to determine the near-field impedance values for the individual electrodes, which are representative of the impedance of tissues in the vicinity of the electrodes. The device then performs or controls various device functions based on the near-field values, such as analyzing selected near-field values to detect heart failure or pulmonary edema.