Table 10. Prediction error comparison for Barrett Universal II, SVM-RM and MLNN-EM for all axial lengths.

Eyes within PE (%)	Barrett Universal II	SVM-RM	MLNN-EM
±0.25	43.5–60.0	48.1	48.5
±0.50	72.3–80.6	82.7	82.3
±1.00	94.5–99.7	97.7	97.7

Note:

Prediction error (PE).