MAIA and Humphrey Perimetry Differ in Their Estimation of Homonymous Visual Field Defects.

Purpose: To assess whether the Humphrey Visual Field Analyzer (HFA) and the Macular Integrity Assessment instrument (MAIA) provide equivalent estimates of visual deficit size, severity, and progression in cortically blinded participants.

Methods: Reliable, monocular 10-2 HFA and MAIA fields were collected at baseline, and after a blind-field training intervention (n = 54) or no intervention (n = 6) in adult participants with occipital strokes. Binocular HFA and MAIA mean sensitivities (MS) were first computed, before creating binocular maps of visual sensitivity for each perimetry system to calculate deficit areas, using a unitary, published, less than 10 dB criterion to define blindness. We contrasted HFA/MAIA MS and deficit area at individual study visits, together with change in these measures between visits.

Results: At individual visits, MS and deficit areas were well-correlated, but there were systematic differences between machines, greater for the intact than impaired hemifields, with the MAIA overestimating areas of visual deficit relative to the HFA. Between visits, the two perimeters' assessment of change in MS was correlated, but change in the deficit area was not, despite good test reliability.

Conclusions: Reliable HFA and MAIA tests produce well-correlated but systematically offset estimates of visual sensitivity and deficit area in patients with homonymous field defects, which can impact assessment of progression.

Translational relevance: Cortically blinded patients exhibited comparable changes in sensitivity (spontaneous and training induced) when assessed using Humphrey and fundus-controlled MAIA perimetry. However, the photopic vs. mesopic nature of these tests impacted participant performance and suggests that custom, machine-specific criteria are needed to comparably define blindness on both systems.