We thank Drs Oishi and Holz for their interest in our concentric rings method that we have proposed in our article. We agree with their comments and applaud the innovative suggestion in improving the concentric rings method using elliptical rings with uneven intervals. In our paper, we acknowledged the distortion applied when projecting a 3-dimensional image to a 2-dimensional image and stressed the importance of correcting these errors prior to the use of the concentric rings. We also appreciate the work by Oishi and associates in identifying a horizontal stretch in Optos 200TX (Optos Plc, Dumfermline, Scotland), and therefore elliptical rings with uneven intervals would be an interesting solution in raw images.

However, we feel that owing to the image aberrations present in ultra-widefield imaging, either from a projection error, from horizontal stretch, or even from the contribution of varying axial lengths, creating a grading method with specific intervals as in the model eye by Oishi and associates will limit its use to a specific axial length, for example 24 mm, as used in the model eye. Furthermore, as rightly mentioned by Oishi and associates, elliptical rings with uneven intervals are only for use in an original image without gaze steering. Steering protocols are increasingly being used and achieve a wider field, but images will need to be corrected for a proper montage to be obtained owing to the distortions incorporated in the periphery in a steered image. Therefore, elliptical rings will not be applicable to steered images.

We appreciate that at the time of the original study, the correction software by Optos was not widely available. However, currently the latest Optos software corrects for this. Therefore, with a corrected image, the concentric rings will be a suitable option in quantifying peripheral nonperfusion in ultra-widefield images.