Visualizing the dynamic change of Ocular Response Analyzer waveform using Variational Autoencoder in association with the peripapillary retinal arteries angle

• Published in: Scientific reports Vol. 10; no. 1; pp. 6592 - 8
• Main Authors: Asano, Shotaro, Asaoka, Ryo, Yamashita, Takehiro, Aoki, Shuichiro, Matsuura, Masato, Fujino, Yuri, Murata, Hiroshi, Nakakura, Shunsuke, Nakao, Yoshitaka, Kiuchi, Yoshiaki
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 20.04.2020; Nature Publishing Group; Nature Portfolio
• Subjects: 692/53/2423; 692/699/3161/3174; Adult; Aged; Aged, 80 and over; Arteries; Cornea; Cornea - blood supply; Cornea - diagnostic imaging; Cornea - physiopathology; Deep learning; Diagnostic Techniques, Ophthalmological - trends; Female; Glaucoma, Open-Angle - diagnosis; Glaucoma, Open-Angle - diagnostic imaging; Glaucoma, Open-Angle - physiopathology; Humanities and Social Sciences; Humans; Intraocular Pressure - physiology; Male; Middle Aged; multidisciplinary; Retina; Retina - diagnostic imaging; Retina - physiopathology; Retinal Artery - diagnostic imaging; Retinal Artery - physiopathology; Science; Science (multidisciplinary); Tonometry, Ocular - methods; Variance analysis
• ISSN: 2045-2322; 2045-2322
• DOI: 10.1038/s41598-020-63601-8

The aim of the current study is to identify possible new Ocular Response Analyzer (ORA) waveform parameters related to changes of retinal structure/deformation, as measured by the peripapillary retinal arteries angle (PRAA), using a generative deep learning method of variational autoencoder (VAE). Fifty-four eyes of 52 subjects were enrolled. The PRAA was calculated from fundus photographs and was used to train a VAE model. By analyzing the ORA waveform reconstructed (noise filtered) using VAE, a novel ORA waveform parameter (Monot1-2), was introduced, representing the change in monotonicity between the first and second applanation peak of the waveform. The variables mostly related to the PRAA were identified from a set of 41 variables including age, axial length (AL), keratometry, ORA corneal hysteresis, ORA corneal resistant factor, 35 well established ORA waveform parameters, and Monot1-2, using a model selection method based on the second-order bias-corrected Akaike information criterion. The optimal model for PRAA was the AL and six ORA waveform parameters, including Monot1-2. This optimal model was significantly better than the model without Monot1-2 (p = 0.0031, ANOVA). The current study suggested the value of a generative deep learning approach in discovering new useful parameters that may have clinical relevance.