Intraoperative optical imaging (IOI) is a localization method for functional areas of the human brain cortex during neurosurgical procedures. The aim of the current work was to develop of a new analysis technique for the computation of two-dimensional IOI activity maps that is suited especially for use in clinical routine. The new analysis technique includes a stimulation scheme that comprises 30-s rest and 30-s stimulation conditions, in connection with pixelwise spectral power analysis for activity map calculation. A software phantom was used for verification of the implemented algorithms as well as for the comparison with the commonly used relative difference imaging method. Furthermore, the analysis technique was tested using intraoperative measurements on eight patients. The comparison with the relative difference algorithm revealed an averaged improvement of the signal-to-noise ratio between 95% and 130% for activity maps computed from intraoperatively acquired patient datasets. The results show that the new imaging technique improves the activity map quality of IOI especially under difficult intraoperative imaging conditions and is therefore especially suited for use in clinical routine.