Geographic Atrophy Area Measurements Vary by FAF Device, Wavelength

Today’s eyecare providers use a variety of imaging platforms to assess lesion area and enlargement in patients with geographic atrophy (GA), including Heidelberg Spectralis and widefield systems like Optos California and Zeiss Clarus 700. Because these devices differ in optics, wavelength and capture methods, researchers performed a systematic cross‑platform comparison to determine how these differences influence GA boundary detection and area measurements. They found that while reproducibility was strong within each system, there was evident variability between devices. A study comparing three FAF imaging platforms—Heidelberg Spectralis, Optos California and Zeiss Clarus 700—for assessing GA lesions observed systematic, device- and wavelength-dependent differences, with Clarus Green and Spectralis manual planimetry measuring larger areas and Optos California smaller ones. Photo: Bogost J, et al. Ophthalmol Sci. April 2026. Click image to enlarge. Included in the study were 61 eyes of 39 participants aged 50 years or older with GA secondary to age-related macular degeneration. Participants were imaged with each of the three FAF systems on the same day in a fixed order: Optos California (532nm green excitation, 200° ultra-widefield), Zeiss Clarus 700 (blue 435nm to 500nm and green 500nm to 585nm channels, 133° widefield) and Heidelberg Spectralis (488nm blue excitation, 30° field). Manual planimetry was used to segment GA for all modalities. Spectralis images were also segmented semiautomatically using RegionFinder, which was prespecified as the reference standard. The main outcome measures were the mean GA area by device and overall agreement among systems.Although all three platforms demonstrated high intra-modality reproducibility and excellent intergrader agreement, GA area estimates differed systematically across devices and wavelengths. On average, GA measured larger with Clarus Green and Spectralis manual planimetry (0.49mm² and 0.34mm², respectively), the former of which identified a modestly higher proportion of eyes with subfoveal GA than the other modalities. In contrast, Optos California measured smaller areas (-0.48mm²) and detected reticular pseudodrusen less frequently. Clarus Blue was excluded from the primary analysis, as nearly a quarter of images from that modality were ungradable due to poor quality. The researchers surmise this is likely because blue-light autofluorescence is more susceptible to ocular media scatter and macular pigment absorption.“Among the alternative imaging platforms evaluated, Clarus Green performed most similarly to Heidelberg manual planimetry in both GA area measurements and visibility of clinically relevant features,” the authors wrote in their paper. “In addition to its high agreement, Clarus Green offers several practical advantages,” they explained, which include “a wider field of view compared to Heidelberg’s 30°, improved visibility of the fovea due to reduced masking by central luteal pigment and better patient comfort owing to less intense illumination.”The researchers concluded that the observed inter-modality differences in GA area measurement “are not insignificant and may influence clinical trial eligibility, endpoint interpretation and progression tracking.” Until device-agnostic algorithms are created to reduce these biases, the authors argued, “Repeated use with the same device [is] recommended for clinical trials.”Click here for the journal source. Bogost J, Saunders T, Heathcote J, et al. Comparison of Clarus, Optos and Heidelberg systems for geographic atrophy area measurements (COCO GA). Ophthalmol Sci. April 2026. [Epub ahead of print]. This article was developed by the editorial staff in conjunction with experts in the field. In the process, AI may have been among the editorial tools used to meet the goals of human editors, who approved all content.