Introduction Susceptibility Weighted Imaging (SWI) is an MR method that employs both magnitude and phase information [1]. The increasing use of phased array coils due to the improved SNR and the desire to reduce acquisition times by parallel imaging techniques such as GRAPPA [2] calls for methods for combining the phase images of the coil elements of a phased array. In case of SWI a common technique for the combination of the phase images is to employ homodyne detection [3] which corrects phase images for phase variations with low spatial frequencies, such as those caused by the inhomogeneous coil sensitivities, and to combine the resulting phase images using a weighted sum [4]. However, homodyne filtering may cause artifacts in areas of strong field inhomogeneities. These artifacts can be avoided if a combination of phase unwrapping and correction for phase variations of low spatial frequencies is used [5,6]. In this case the phase images have to be combined correctly before unwrapping. Incorrectly combined phase images can be affected by singularities in areas of sufficient signal. Such singularities impede phase unwrapping. On Siemens scanners the combined phase image is either calculated using Adaptive Combine [7] or a complex summation. Both can result in incorrect phase images affected by singularities. Methods We implemented a uniform sensitivity reconstruction [8] as a functor in the Siemens ICE framework. This functor was integrated into the standard reconstruction functor pipeline after the GRAPPA decorator. The functor relies on sensitivity maps which are automatically computed from a low resolution prescan using the body coil and one with the respective phased array. The maps are computed according to a method proposed by Pruessmann et al [9]. For the prescan a 2D gradient echo