Magnetic resonance imaging (MRI) has emerged as a powerful tool in medical diagnosis and research. Although high spatial resolution images are essential in medical diagnosis and image analysis, high temporal resolution is equally important in applications of dynamic contrast-enhanced MRI or functional brain MRI. In particular, in breast MRI the ability to differentiate between benign and malignant lesions depends, in part, on the temporal resolution of the dynamic image acquisition. New applications of MRI such as multi-feature analysis of image time series data and full 3D functional MRI or event-related functional MRI require high spatial and high temporal resolution for accurate image analysis on a voxel-by-voxel basis. Currently available partial Fourier reconstruction techniques. which effectively improve the time resolution, suffer from a reduced signal to noise ratio in the reconstructed image, a decrease in spatial resolution or reconstruction artefacts, making numerical image analysis difficult. In this work we present an image reconstruction algorithm based on image recovery theory which effectively doubles the temporal resolution and results in an image quality sufficient for further numerical analysis. The developed algorithm requires a full Fourier space acquisition of a pre-contrast or baseline image prior to the reconstruction procedure of the time series partial Fourier data.