Author response: Visual discrimination training improves Humphrey perimetry in chronic cortically induced blindness

We thank Drs. Wall and Schiefer for the comment on our article.¹ While we agree that Humphrey visual fields (HVFs) are not perfect tests, they remain the clinical gold standard for measuring visual defects in hemianopia. There are several problems with Drs. Wall and Schiefer's statements. First, a shift in attention does not automatically cause a shift in fixation, as addressed in the extensive literature on covert visual attention,² and whose manipulation we are pursuing actively in NIH-funded work to enhance visual recovery in hemianopic fields. Second, while 24-2 HVFs have a resolution of 6°, the 10-2 HVFs have a resolution of 2°. As stated in our article,¹ we combined and interpolated these 2 tests and the 2 eyes algorithmically, creating maps with a resolution of 0.1° squared. Because these maps were formed from the averaging and interpolation of 4 separate fields, any small spontaneous fixation shift would have had to occur consistently 4 separate times for locations within the inner 10°, and twice for locations greater than 10° from fixation. Moreover, we excluded patients with abnormal false-positive, false-negative, and fixation loss rates and our 10-2 short-term fluctuations were approximately 2 dB.¹ Ultimately, our inclusion of untrained patients provided the best evidence that recovery seen in trained patients was not due to test-retest variability or learning to perform better on the test. Only participants with improved HVFs had improved psychophysical performance under strict fixation-controlled conditions enforced binocularly using an Eyelink1000 eye tracker with superior spatial and temporal resolution.¹ It was precisely to address the types of criticisms raised by Drs. Wall and Schiefer that we proposed a new approach to collecting and analyzing HVFs, while rigorously controlling for drifts in fixation, whether attention-induced or otherwise.