Abstract
The hypothesis that surface draining veins distort measurements in human visual area V4 (hV4) offers an explanation for why retinotopic maps measured in this region often appear to contain only an incomplete hemifield (Winawer et al., 2010). Puckett et al. (2014) indicated that voxels contaminated by venous artefact display inverted responses to visual stimulation. Here, we aimed to quantify the relationship between venous eclipses and hV4 map coverage by assessing the proximity and behaviour of voxels in visual cortex contaminated by venous artefact. Subjects (N=11) viewed bowtie, ring, drifting bar and full field flash stimuli. A 0.75mm3 T1 anatomical image and 1.5mm3 functional EPIs were acquired over 1.5hrs. Visual areas were identified using pRF polar angle maps, correlation analyses were used to identify inverted voxels, and mean EPI luminance maps were used to identify venous eclipses. Cortical surfaces were generated using mrMesh and Caret 5 and functional analyses were conducted using mrVista. No consistent relationship was found between the presence or absence of a venous eclipse and incomplete maps of hV4. Furthermore, there was no statistically significant difference in the number of inverted voxels present across visual areas V1 to hV4. Mean maps 2.5mm above the grey/white boundary accurately reveal locations of larger veins, showing for the first time exactly where venous artefact would be expected to impact retinotopic maps. Venous eclipses appear in conjunction with incomplete maps of hV4, however cannot explain every instance. Complete maps of hV4 were sometimes present despite a strong venous eclipse in immediate proximity. It remains unclear whether venous artefact causes increased numbers of inverted voxels in hV4, however with more accurate images of the venous eclipse, it is possible that future analyses, including across several laminae, will further disambiguate problems affecting our ability to accurately map hV4.
Meeting abstract presented at VSS 2018