To fulfill the battery requirements of an electric vehicle in terms of power, driving range and safety an accurate online estimation of the cell states is essential. In this paper two data-based approaches, which continuously estimate the resistance of each cell with low computational effort and memory capacity needed, are presented. The long- and short-term changes of the estimated resistance distribution are used to give insights into temperature and SoH gradients between the cells of a battery. The algorithms are validated by measurements of a battery, consisting of two modules each containing 50 automotive lithium-ion pouch cells. Decreasing cell temperatures close to the module edges are detected by the algorithms, which is validated by the use of temperature sensors.