Table 2.

Agreement of Task-Specific Labels Across Training, Validation and Test Sets, Measured via Cohen's Kappa Statistic, Which Is Essentially a Number Between −1 and 1. Whereas 1 Indicates a Full Agreement, Lower Scores Mean Less Agreement. Negative Scores Indicate Disagreement

	Training			Validation			Test
	Subretinal Fluid	Intraretinal Fluid	Active nAMD	Subretinal Fluid	Intraretinal Fluid	Active nAMD	Subretinal Fluid	Intraretinal Fluid	Active nAMD
Subretinal fluid	–	−0.02	0.79	–	0.26	0.75	–	−0.02	0.59
Intraretinal fluid	−0.02	–	0.37	0.26	–	0.65	−0.02	–	0.57