In Journal of glaucoma
PRCIS : The offline AI on a smartphone-based fundus camera shows good agreement and correlation with the vCDR from the SD-OCT and manual grading by experts.
PURPOSE : To assess the agreement of vertical cup-to-disc ratio (vCDR) measured by a new artificial intelligence (AI) software from optic disc images obtained using a validated smartphone-based imaging device, with spectral-domain optical coherence tomography (SD-OCT) vCDR measurements, and manual grading by experts on a stereoscopic fundus camera.
METHODS : In a prospective, cross-sectional study, participants >18 years (Glaucoma and normal) underwent a dilated fundus evaluation followed by optic disc imaging including a 42o monoscopic disc-centered image (Remidio NM FOP-10), a 30o stereoscopic disc-centered image (Kowa nonmyd WX-3D desktop fundus camera), and disc analysis (Cirrus SD-OCT). Remidio FOP images were analysed for vCDR using the new AI software and Kowa stereoscopic images were manually graded by three fellowship-trained glaucoma specialists.
RESULTS : We included 473 eyes of 244 participants. The vCDR values from the new AI software showed strong agreement with SD-OCT measurements (95% limits of agreement (LoA)=-0.13 to 0.16). The agreement with SD-OCT was marginally better in eyes with higher vCDR (95% LoA=-0.15 to 0.12 for vCDR>0.8). Interclass correlation coefficient (ICC) was 0.90 (95% CI:0.88-0.91). The vCDR values from AI software showed a good correlation with the manual segmentation by experts (ICC=0.89, 95%CI:0.87-0.91) on stereoscopic images (95% LoA=-0.18 to 0.11) with agreement better for eyes with vCDR>0.8 (LoA=-0.12 to 0.08).
CONCLUSION : The new AI software vCDR measurements had an excellent agreement and correlation with the SD-OCT and manual grading. The ability of the Medios AI to work offline, without requiring cloud-based inferencing, is an added advantage.
Shroff Sujani, Rao Divya Parthasarathy, Savoy Florian M, Shruthi S, Hsu Chao-Kai, Pradhan Zia Sultan, Jayasree P, Sivaraman Anand, Sengupta Sabyasachi, Shetty Rohit, Rao Harsha L
2022-Nov-28
