To optimize spatial resolution in single photon emission tomography (SPET), it is essential to minimize the radius of rotation. In brain studies, different methods have been used to avoid shoulder interference when the radius of rotation is minimized: rectangular fields of view, modifications to the shielding around circular detectors and fan or cone beam collimators. However, few single-head systems can adopt these developments, particularly older cameras. A non-standard image acquisition method to reduce the radius of rotation in brain SPET with a single-head gamma camera is presented. The method applies a defined transformation to the original acquired images, maintaining the whole of the brain inside the field of view without shoulder interference and meeting the condition: pixel size < or = FWHM/3. With this method, it is possible to reduce the radius of rotation to 16 cm and to obtain a transaxial spatial resolution of 15.98 mm, which is 3.5 mm less than with the standard method used in our laboratory. This procedure was implemented for a Siemens Gammasonics ZLC 3700 gamma camera and has been validated in single-slice brain phantom studies. The method has the advantage of not requiring any complex or costly hardware.