Fluorescein angiography (FA) has been used to visualize the retinal blood vessels and perfusion thereby enhancing the knowledge of the pathophysiology underlying retinal diseases (Pineles and Arnold, 2012; Desmettre et al., 2000). It is an invasive imaging technique which can be used for diagnostics of retinal and choroidal diseases as well as for research (Talks, J. et al., 2007; Oto, S. et al., 2002; Pauleikhoff, D. et al., 1999). Another method for examining the retinal microcirculation is retinal oximetry, which is a non-invasive imaging technique that utilizes spectrophotometry and allows reproducible calculation of retinal blood vessel oxygen saturation and vessel width (Klefter et al., 2015). In brief, oxygen saturations are typically measured in the peripapillary vessels and are estimated based on the intensity of reflected light from retinal vessels and surrounding tissue. The underlying principle utilizes the difference in the light absorption spectra of oxyhemoglobin and deoxyhemoglobin in the blood (Schweitzer et al., 1999, 2001; Hardarson et al., 2006; Vosborg et al., 2020). Retinal oximetry has been validated with both hypoxia and hyperoxia as well as calibrated towards systemic arterial saturation through extensive studies (Eliasdottir, 2018). The Oxymap retinal oximetry apparatus provides automated vessel detection and estimates oxygen saturations in all vessels thus detected in the image. Manual tracing allows for particular vessel segments to be selected for analysis. This makes the analysis adaptable and allows for customization to different clinical research questions. Differentiating between arterial and venous segments is fundamental. Traditionally, vessel selection for retinal oximetry analysis is done based on morphological characteristics and supported by an overlay on the fundus image showing the estimated blood oxygen saturations. To examine the validity of this approach, we compared it with vessel selection supported by fluorescein angiography.

This validation study was performed in patients with optic disc drusen (ODD). This is an optic neuropathy where compromised blood flow to the anterior part of the optic nerve head is believed to be an essential part of the pathophysiology underlying the visual dysfunction (Hamann et al., 2018; Palmer et al., 2018). Hence, retinal oximetry has the potential to characterize key aspects of the pathophysiology of ODD which was therefore chosen as the model disease for the present investigation. We aimed to determine if the use of fluorescein angiography guidance would lead to different retinal oximetry results when compared with the traditional method of manual vessel selection. The results could be valuable in future analyses of patients with ODD as well as other retinal or optic nerve diseases.

A previous study (Mustonen and Niemien, 1982) proposed that FA is a valid method to examine abnormal vascular patterns that could be associated with ODD. Additionally, earlier studies have found a prolonged filling of the peripapillary choriocapillaris in patients with ODD and suggest that this finding is an indication of an infarct in the choroid or impaired blood flow (de Laey, 2014; Pauleikhoff, D. et al., 1999; Erkkilä, 1976). Several other studies have shown a correlation between perfusion velocities and oxygen saturation suggesting that the saturation can be used as an indicator of the blood flow of the retina (Stefánsson et al., 2019; Jeppesen and Bek, 2019; Blindbæk et al., 2020). In glaucoma patients, oximetry findings have been assessed to further support the rationale of using oximetry in examination of patients with optic neuropathies (Vandewalle et al, 2014, Shughoury et al., 2020).