2.4. Outcome measures

Diagnostic evaluations were carried out at three time points: a baseline assessment prior to the intervention, a post-treatment assessment two to five days after the intervention, and a follow-up assessment at eight weeks (Fig. 1). Examinations included SAP, HRP, visual acuity, and reading performance in all centers. Additionally, contrast sensitivity was performed in Experiments 1 and 3 and dynamic visual acuity (DVA) in Experiment 1.

The primary outcome was change in VFD determined with monocular SAP and binocular computer-based HRP (Kasten et al., 1998). The former measures near-threshold light detection in the 30° visual field, whereas the latter assesses super-threshold visual detection of 12° central vision. In HRP, visual field change was defined by a percentage of correctly-detected stimuli (detection accuracy, DA) and in SAP by mean stimulus detection sensitivity (MS), comparing baseline with post treatment and follow-up. Secondary outcomes included additional variables derived from HRP, SAP, and other functional visual tests.

Static monocular visual field perimetry was obtained with a near-threshold Static Automated Perimetry (Oculus Twinfield, Lynnwood, WA in Experiment 1, Octopus 900, Haag-Streit Diagnostics in Experiment 2, and Humphrey Feld Analyzer II-i 750i in Experiment 3). We determined threshold values at 59–76 positions of the 30° visual field, where target stimuli (size: III/4 mm², color: white, duration: 0.2 s) were presented on a background with constant luminance of 10 cd/m². MS of all test positions and foveal sensitivity were analyzed. Additionally, in Experiment 1, the sizes of the absolute and relative defect and the intact field were defined as the fraction of test positions detected at none, some, and all of the test times during a session.

In the computer-based HRP, static supra-threshold white dots lasting 150 ms were presented in a randomized sequence at 400 positions of a 21×21 grid on a dark screen; the location and number of correctly detected stimuli were registered as described previously (Kasten et al., 1998). Each patient performed three binocular tests that were superimposed to define intact, relative defect, and absolute defect areas of the visual field. Detection accuracy, reaction times, and false positive responses were recorded.

To rule out that eye movements influenced HRP results, fixation was monitored with an online eye tracker (Tobii X2-60 Eye Tracker, Tobii Technology AB, Sweden) in Experiments 1 and 2. Test points during which a subject’s gaze deviated outside the predefined area of 2° were disregarded and repeated at the end of the test round. After three consecutive failures, the test was paused, and the subject was encouraged to maintain fixation. Calibration was run to achieve a mean fixation error < 1°. In Experiment 3, the subjects underwent microperimetry to ensure sufficient fixation. Additionally, in all experimental arms, during 10% of the trials, the fixation point changed color isoluminantly to which subjects had to react; the response rate was considered a surrogate of fixation accuracy. In SAP, fixation monitoring depended on the perimetry used; it included either automatic fixation control or manual control with recorded fixation losses, and visual monitoring by a researcher performing the measurement.

Other measures of visual function included near visual acuity (Oculus® in Experiment 1 and 2 and MNREAD acuity chart in Experiment 3) and reading performance. Reading was examined with international reading speed test (IResT) (Trauzettel-Klosinski & Dietz, 2012), validated for German, Finnish, and Italian languages. The MARS contrast sensitivity test (Dougherty et al., 2005) and computer-assisted DVA test (Wist et al., 1998) were assessed in a subset of experiments as described above.