Treffer: Comparison of machine learning-based algorithms using corneal asymmetry vs. single-metric parameters for keratoconus detection.
Romero-Jimenez M, Santodomingo-Rubido J et al (2010) Keratoconus: a review. Cont Lens Anterior Eye 33:157–66 (quiz 205). (PMID: 10.1016/j.clae.2010.04.00620537579)
Mihaltz K, Kovacs I, Takacs A et al (2009) Evaluation of keratometric, pachymetric, and elevation parameters of keratoconic corneas with pentacam. Cornea 28:976–980. (PMID: 10.1097/ICO.0b013e31819e34de19724217)
Prakash G, Suhail M, Srivastava D (2016) Predictive analysis between topographic, pachymetric and wavefront parameters in keratoconus, suspects and normal eyes: creating unified equations to evaluate keratoconus. Curr Eye Res 41:334–342. (PMID: 25803133)
Toprak I, Yaylali V, Yildirim C (2015) A combination of topographic and pachymetric parameters in keratoconus diagnosis. Cont Lens Anterior Eye 38:357–362. (PMID: 10.1016/j.clae.2015.04.00125936634)
Arbelaez MC, Versaci F, Vestri G et al (2012) Use of a support vector machine for keratoconus and subclinical keratoconus detection by topographic and tomographic data. Ophthalmology 119:2231–2238. (PMID: 10.1016/j.ophtha.2012.06.00522892148)
Belin MW, Keratoconus DJK (2016) The ABCD grading system. Klin Monbl Augenheilkd 233:701–707. (PMID: 10.1055/s-0042-10062626789119)
Duncan JK, Belin MW, Borgstrom M (2016) Assessing progression of keratoconus: novel tomographic determinants. Eye Vis (Lond) 3:6. (PMID: 10.1186/s40662-016-0038-626973847)
Mahmoud AM, Roberts CJ, Lembach RG et al (2008) CLMI: the cone location and magnitude index. Cornea 27:480–487. (PMID: 10.1097/ICO.0b013e31816485d3184348542861357)
Sedghipour MR, Sadigh AL, Motlagh BF (2012) Revisiting corneal topography for the diagnosis of keratoconus: use of Rabinowitz’s KISA% index. Clin Ophthalmol 6:181–184. (PMID: 223319753273406)
de Sanctis U, Loiacono C, Richiardi L et al (2008) Sensitivity and specificity of posterior corneal elevation measured by Pentacam in discriminating keratoconus/subclinical keratoconus. Ophthalmology 115:1534–1539. (PMID: 10.1016/j.ophtha.2008.02.02018405974)
Ruiz Hidalgo I, Rodriguez P et al (2016) Evaluation of a machine-learning classifier for keratoconus detection based on Scheimpflug tomography. Cornea 35:827–832. (PMID: 10.1097/ICO.000000000000083427055215)
Issarti I, Consejo A, Jimenez-Garcia M et al (2019) Computer aided diagnosis for suspect keratoconus detection. Comput Biol Med 109:33–42. (PMID: 10.1016/j.compbiomed.2019.04.02431035069)
Tong Y, Lu W, Yu Y et al (2020) Application of machine learning in ophthalmic imaging modalities. Eye Vis (Lond) 7:22. (PMID: 10.1186/s40662-020-00183-632322599)
Yousefi S, Elze T, Pasquale LR et al (2020) Monitoring glaucomatous functional loss using an artificial intelligence-enabled dashboard. Ophthalmology 127:1170–1178. (PMID: 10.1016/j.ophtha.2020.03.00832317176)
Treder M, Lauermann JL, Eter N (2018) Automated detection of exudative age-related macular degeneration in spectral domain optical coherence tomography using deep learning. Graefes Arch Clin Exp Ophthalmol 256:259–265. (PMID: 10.1007/s00417-017-3850-329159541)
Herber R, Pillunat LE, Raiskup F (2021) Development of a classification system based on corneal biomechanical properties using artificial intelligence predicting keratoconus severity. Eye Vis (Lond) 8(1):21. (PMID: 10.1186/s40662-021-00244-434059127)
Lopes BT, Ramos IC, Salomao MQ et al (2018) Enhanced tomographic assessment to detect corneal ectasia based on artificial intelligence. Am J Ophthalmol 195:223–232. (PMID: 10.1016/j.ajo.2018.08.00530098348)
Smadja D, Touboul D, Cohen A et al (2013) Detection of subclinical keratoconus using an automated decision tree classification. Am J Ophthalmol 56:237-246 e1. (PMID: 10.1016/j.ajo.2013.03.034)
Kasun LLC, Yang Y, Huang G et al (2016) Dimension reduction with extreme learning machine. IEEE Trans Image Process 25:3906–3918. (PMID: 10.1109/TIP.2016.257056927214902)
Weitere Informationen
Purpose: To evaluate the diagnostic performance of three different parameter sets relevant to corneal asymmetry in comparison to conventional parameters including maximum anterior corneal curvature (K <subscript>max</subscript> ) and thinnest corneal thickness for diagnosis of keratoconus.
Methods: In this retrospective case control study, 290 eyes with keratoconus and 847 eyes of normal patients were included in the analyses. Corneal tomography data were acquired from Scheimpflug tomography. The sklearn and FastAI libraries were used in a Python 3 environment to create all machine learning models. The original topography metrics and derived metrics together with the clinical diagnoses were used as the dataset for model training. The data were first split to assign 20% of the data to an isolated test set. The remaining data were then split 80/20 to a training and validation group for model training. Sensitivity and specificity outcomes with standard parameters (K <subscript>max</subscript> , central curvature, and thinnest pachymetry) and ratio of asymmetry across horizontal, apex centered, and flat axis-centered axis of reflection were studied via various machine learning models.
Results: Thinnest corneal pachymetry and K <subscript>max</subscript> were 549.8 ± 34.3 µm and 45.3 ± 1.7 D in normal eyes and 460.5 ± 62.6 µm and 59.3 ± 11.3 D in keratoconic eyes. Use of only corneal asymmetry ratios across all 4 meridians had mean sensitivity of 99.0% and mean specificity of 94.0%, better than utilizing K <subscript>max</subscript> alone or traditional measures combined (K <subscript>max</subscript> , thinnest cornea and inferior-superior asymmetry).
Conclusions: By using the ratio of asymmetry between corneal axes alone, a machine learning model could identify patients with keratoconus in our dataset with satisfactory sensitivity and specificity. Further studies on pooled/larger datasets or more borderline population can help validate or refine these parameters.
(© 2023. The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.)