Mapping of Contralateral Spacein Retinotopic Coordinates by aParietal Cortical Area inHumansM. I. Sereno,1* S. Pitzalis,1† A. Martinez,2‡The internal organization of a higher level visual area in the human parietalcortex was mapped. Functional magnetic resonance images were acquired whilethe polar angle of a peripheral target for a delayed saccade was graduallychanged. A region in the superior parietal cortex showed robust retinotopicmapping of the remembered target angle. The map reversed when the directionof rotation of the remembered targets was reversed and persisted unchangedwhen study participants detected rare target reappearances while maintainingfixation, or when the eccentricity of successive remembered targets was un-predictable. This region may correspond to the lateral intraparietal area inmacaque monkeys.The neocortex of each hemisphere in pri-mates consists of a mosaic of nearly 100visual, auditory, somatosensory, motor, andlimbic areas. To precisely define their bor-ders, studies in nonhuman primates havecombined intensive microelectrode mapping,tracer injections, histological stains, andfunctional analysis of single unit responsesand have often visualized these results onunfolded representations of the cortex (1–5).Functional magnetic resonance imaging(fMRI) in combination with retinotopic map-ping stimuli (6 ) and cortical surface recon-struction (7–9) have made it possible to de-fine the borders of early visual areas withcomparable precision in humans (10–12).However, these methods have so far provedless useful in higher visual areas, which haveinstead been defined almost solely by theirresponses during different cognitive tasks.The borders of many higher visual areas inhumans have yet to be precisely defined.Studies of brain-lesioned humans havelong implicated the parietal cortex in spatialprocessing (13, 14 ). This has been confirmedby recent neuroimaging studies (15–21). Sin-gle-neuron recordings in awake behaving ma-caques suggest that parietal cortical neuronscode and update target locations in retinocen-tric coordinates (22–26 ). It is less clear, how-ever, whether remembered locations in con-tralateral visual space are systematically ar-rayed across subregions of the cortex. Somehigher level areas in nonhuman primates areknown to contain systematic maps of morecomplex stimulus features. For example, de-spite being insensitive to the exact retinalposition of a face, small regions of the infero-temporal cortex appear to code for a sequenceof orientations of the face ( profile, three-quarters, and frontal) at a sequence of nearbycortical locations (27).Study participants (n ⫽ 12) viewed stim-uli projected onto a screen at their chin levelvia a mirror adjusted so that central fixationwas comfortable (28). In the first task, par-ticipants made delayed saccades (Fig. 1, top)(29). A brief peripheral target was presentedwhile participants fixated a central point. Aring of target-sized blinking distractors nearthe target eccentricity then appeared during a3-s delay period while participants main-tained fixation. At fixation dimming (or off-set) (30) and distractor offset, participantsmade a saccade from the fixation point to theremembered target location on a black screen.Then they immediately made a saccade backto the fixation point, which brightened (orreappeared) in preparation for the next pe-ripheral target. The time between successivetarget onsets was 5 s.The angle of the remembered target loca-tion was stepped in a counterclockwise(CCW) direction through 360° so that imag-ing data could be analyzed with the same1Department of Cognitive Science,2Department ofRadiology, University of California, San Diego, La Jolla,CA 92093-0515, USA.*To whom correspondence should be addressed. E-mail: [email protected]†Present address: Laboratory of Functional Neuroim-aging, Fondazione Santa Lucia, Instituto di Ricovero eCura a Carattere Scientifico, Rome, Italy.‡Present address: Sackler Institute, Weill Medical Col-lege of Cornell University, New York, NY, USA.Fig. 1. Phase-encodedstimuli for generatingmaps of rememberedlocation. In the firstthree tasks, the personis fixating centrallywhen a peripheral tar-get briefly appears at12° to 15° of eccen-tricity. A ring of blink-ing distractors appearswhile the person re-members the targetlocation and main-tains fixation for 3 s.In the first and secondtasks (top), the personsaccades to the re-membered locationafter the delay periodand then immediatelysaccades back to thefixation point to pre-pare for the next tar-get, which appearsCCW (or CW ) to thelast. In the third task(middle), participantsmerely detect whenthe target occasional-ly reappears amongthe distractors (it re-appears on averageonce every 12 targets)and make no saccades.The fourth task (bot-tom) is similar to thefirst except that theeccentricity of eachtarget is completely unpredictable (3° to 15°) and the distractors fill the entire field. In every case,the peripheral targets cover 360° of polar angle in 64 s.R EPORTS9 NOVEMBER 2001 VOL 294 SCIENCE www.sciencemag.org1350Fourier-based method used to map polar an-gle in retinotopic visual areas (10). The tar-gets were jittered slightly in eccentricity andpolar angle (2.5° visual angle, both axes) tomake them somewhat less predictable. Onecomplete polar angle cycle took 64 s, so theaverage step in polar angle to the next targetwas about 28°.Three additional versions of the task werethen used to control for artifacts. In the sec-ond task (Fig. 1, top), successive targets ap-peared in a clockwise (CW) instead of aCCW direction. In the third task (Fig. 1,middle), participants had to detect the occa-sional reappearance (P ⫽ 0.08) of CCW-progressing targets among the flashing dis-tractors while maintaining central fixation; inthis task, they made no saccades during theentire scan. The fourth task (Fig. 1, bottom)was like the first except that the eccentricityof the next target was completely unpredict-able, and the distractors filled the field.Each functional session included a seriesof surface-coil echo-planar scans (3 ⫻ 3 ⫻ 4mm voxels, 128 images per slice) focused onoccipital and parietal visual areas and a singlehigh-resolution alignment scan, which wasused to register the functional data with acortical surface reconstruction made fromhead-coil structural scans taken in a separatesession (31).Figure 2 shows the regions in the righthemisphere of one person that were activatedin a