Skip to main content
Fig. 1 | BioMedical Engineering OnLine

Fig. 1

From: Predicting post-operative vault and optimal implantable collamer lens size using machine learning based on various ophthalmic device combinations

Fig. 1

Top five features in best-performing models for vault prediction using data from different devices and combinations. SHAP summary plots for the top five features in vault prediction algorithms with data from different ophthalmic devices and combinations (A–G), and various devices without UBM, Pantacam, and Sirius (H). The higher the SHAP value for each feature, the higher risk of vault increase. SHAP Shapley Additive Explanations; ACD central anterior chamber depth; STS_H horizontal sulcus-to-sulcus; STS_V vertical sulcus-to-sulcus; ICL Implantable Collamer Lens; AL axial length; Kf flat keratometry; Ks steep keratometry; ACV anterior chamber volume; size-STS the difference between ICL size and STS_H; WTW horizontal white-to-white; size-WTW the difference between ICL size and WTW; CLR crystalline lens rise; SE: spherical equivalent; IOP intraocular pressure; UCVA uncorrected distance visual acuity

Back to article page