Research Article

Keratoconus Severity Classification Using Features Selection and Machine Learning Algorithms

Table 1

Summary of previous works in keratoconus classification since 2012.

AuthorsYearMethodDatasetInputsAccuracyFeature selection

Al-Timemy et al. [8]2021SqN, AlN, SfN, MbN2136 imagesN.A92.2% to 94.8%N.A
Kamiya et al. [9]2021CNN3390 imagesN.A78.5%N.A
Jiménez-García et al. [10]2021TDNN743 images6N.AYes
Kuo et al. [11]2020VGG16, InceptionV3, ResNet152354 imagesN.A93.1%, 93.1%, 95.8%N.A
Cao et al. [22]2020RF, SVM, KNN, LR, LDA, LaR, DT, MPAN88 eyes1187%, 86%, 73%, 81%, 81%, 84%, 80%, 52%Yes
Lavric et al. [24]202025 classifiers3151 images862% to 94%SS, FRank
Velázquez-Blázquez et al. [23]2020LR178 eyes573%, Kruskal-Wallis
Lavric and Valentin [12]2019CNN3000 (images)99.33%Yes
Issarti et al. [13]2019FNN851 (images)96.56%NCAFS
Salem and Solodovnikov [14]2019RF500N.A76%Yes
Hallett et al. [15]2019BNN1242973% (supervised) 80% (unsupervised)PCA
Luna et al. [16]2019BNN6016100%N.A
Kamiya et al. [17]2019CNN543 (image)99.1%N.A
Yousefi et al. [5]2018UnML3156420N.APCA NonLinear_tSNE
Hidalgo et al. [18]2017SVM1312592.6% to 98%N.A
Ali et al. [21]2017SVM401290%N.A
Smadja et al. [19]2013DT37255N.AN.A
Arbelaez et al. [20]2012SVM3502798.2%N.A