Abstract: The method and apparatus furnish difference images from a video signal which contains a sequence of images that are obtained from a patient under examination. First a first of the images contained in the video signal is recorded in a first storage device. Thereby a so-called primary mask is formed. Then, a subsequent second of the images contained in the video signal is recorded in a second storage device. Subsequently, a third of these images is recorded in a third storage device. From here on, images contained in the video signal following the third image are recorded alternatingly in the second and the third storage device. At a certain point of time, a signal for starting data acquisition is provided. Simultaneously or thereafter, a so-called secondary mask is selected. The secondary mask is an image completely recorded in the second or third storage device shortly prior to the generation of the data acquisition signal.