Cuantificación automatizada de la no perfusión en la retinopatía diabética mediante angiografía de tomografía de...

To evaluate the performance of a customized deep learning algorithm for automated segmentation of nonperfusion area (NPA) on ultra-widefield swept-source OCTA (UWF SS-OCTA) and its utility in diabetic retinopathy (DR) severity assessment.

Cross-sectional study.

A total of 180 eyes from 122 participants representing all grades of DR severity.

We developed a convolutional neural network based on a multiscale U-Net backbone with squeeze-and-excitation attention for segmentation of NPAs on en face SS-OCTA all-retinal-layer images from 3 scan patterns: 6 × 6 mm, 12 × 12 mm, and 29 × 24 mm. Ground-truth annotations of NPAs and nongradable area (NGA) on en face OCTA images were generated by 2 independent graders and adjudicated by a vitreoretinal specialist. A corresponding en face structural OCT image was incorporated to distinguish true NPAs from shadow artifacts. Segmentation outputs included NPA, NGA, and shadow artifacts. Pixel-level accuracy was assessed with the F1 score. Nonperfusion index (NPI) was defined as NPA/gradable area. The level of agreement between human-labeled and algorithm-predicted NPI was analyzed using Bland-Altman analysis.

Algorithm F1 score and NPI.

The algorithm for NPA segmentation achieved a mean F1 score of 0.82 ± 0.01 in 6 × 6 mm, 0.84 ± 0.03 in 12 × 12 mm, and 0.83 ± 0.02 in 29 × 24 mm, with no significant difference across fields of view (P = 0.12). Algorithm-derived NPI strongly agreed with expert grading (intraclass correlation coefficient >0.979). Both human- and algorithm-derived NPI increased progressively with increased DR severity in all scan patterns demonstrated by the Kruskal-Wallis test (6 × 6 mm: human: P = 0.02; algorithm: P = 0.03; 12 × 12 mm: algorithm P < 0.001; human P < 0.001; 29 × 24 mm: algorithm: P < 0.001; human: P < 0.001) with the largest magnitude of increase in 29 × 24 mm scans. The algorithm for foveal avascular zone segmentation also achieved a mean F1 score of 0.88 ± 0.05 for 6 × 6 mm images and 0.85 ± 0.05 for 12 × 12 mm images.

This deep learning algorithm was validated on single-scan UWF SS-OCTA for automated NPA segmentation and quantification. It demonstrates high accuracy and scalability across multiple scan sizes, supporting its potential integration into objective DR OCTA biomarker analysis.

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.