Cochlear implants (CIs) are effective auditory neuroprostheses, but performance varies across patients. One factor contributing to this variability is the spread of electrical current within the cochlea, which reduces spectral resolution and limits speech understanding. Far-field resistance, which can be measured directly, reflects current flow through perilymph and surrounding tissue and may be influenced by patient-specific anatomy. We developed an image-based model to automatically predict far-field resistance profiles using patient-specific cochlear geometries derived from preoperative computed tomography (CT) scans. In a retrospective analysis of 36 CI cases, surface meshes of the scala tympani and scala vestibuli were automatically segmented. Electrode positions were extracted from postoperative CT scans for validation purposes. Far-field resistance was estimated by computing cumulative resistances along longitudinal pathways within the scalae. The model systematically overestimated resistances by up to a factor of 5.2 depending on electrode insertion depth. After systematic deviation correction, high agreement with measured values was achieved in 33 of 36 subjects (root mean square error less than 0.15 kΩ, Spearman ρ=0.83), while outliers suggested atypical current spread due to electrode positioning, tissue contact, or alternative conductive pathways. The image-based model captured subject-specific resistance trends and apical electrode variability better than a group-average reference, though basal predictions were affected by near-field contributions and segmentation uncertainties. These results demonstrate that individualized CT-based modeling can approximate far-field resistance profiles and could enable further research on the interpretation of intracochlear current flow. Our approach could use routinely acquired preoperative imaging to complement the measured resistance, enabling the interpretation of inter-subject variability and the identification of atypical profiles potentially related to electrode positioning or tissue changes. The approach could be integrated into clinical workflows and extended with advanced modeling techniques to support surgical planning, postoperative monitoring, and device programming, linking anatomical imaging with functional impedance assessment.
