How "Micro" Is Microperimetry? Characterizing the Effect of Fundus Tracking on the Psychometric Function

The purpose of this study was to quantify whether fundus tracking in microperimetry improves psychometric parameter estimation (as an in vivo demonstration of improved stimulus-delivery precision), and to derive a psychometrically grounded criterion intensity for suprathreshold (defect-mapping) microperimetry.

Twenty-five healthy volunteers underwent Macular Integrity Assessment (MAIA)2 microperimetry at five loci referenced to Bruch's membrane opening (BMO): three outside and two inside the blind spot. At each location, frequency-of-seeing (FoS) functions were measured with a method of constant stimuli in four blocks (2 had tracking on and 2 had tracking off). FoS data were fit with a cumulative-Gaussian psychometric function estimating sensitivity, spread, guess, and lapse rate. Tracking effects on m, denoting the sensitivity corresponding to the midpoint of the psychometric function, and s, the spread parameter of the underlying Gaussian, were tested per locus using mixed-effects models. The optimal criterion intensity to separate "non-seeing" from "seeing" loci was derived from FoS fits using posterior predictive simulations.

Tracking had little effect on sensitivity estimates at loci outside the blind spot, but produced lower sensitivity estimates within the blind spot - indicating fewer spurious detections at non-seeing loci (posterior median difference, 95% credible intervals [CrIs] of -1.78 decibel [dB] 95% CrI = -2.58 to -0.94 at locus 4, and -0.75 dB, 95% CrI = -1.71 to 0.15 at locus 5). Tracking was also associated with a smaller spread (steeper FoS curves) at loci 1 to 3, with posterior median differences of -0.13 dB, 95% CrIs = -0.28 to 0.02, -0.27 dB, 95% CrI = -0.43 to -0.12, and -0.25 dB, 95% CrI = -0.42 to -0.07, respectively. Without tracking, false-positive responses were more frequent when fixation shifts displaced stimuli toward the "seeing" retina. Simulation-based analysis identified 13 dB as the nominally optimal criterion intensity for suprathreshold microperimetry (Youden index = 0.74, 95% CrI = 0.70-0.77); a 10 dB criterion yielded comparable discrimination (Youden index = 0.72, 95% CrI = 0.69-0.75).

Even in healthy volunteers with stable fixation, fundus tracking measurably lowered apparent sensitivity at "non-seeing" loci (i.e., reduced spurious responses inside the blind spot) and steepened FoS curves in the "seeing" retina. A criterion intensity in the range of 10 to 16 dB is a defensible choice for separating "seeing" and "non-seeing" retina in suprathreshold (defect-mapping) perimetry paradigms.