Table 7.

Concordance between a software-identified match within the top ten closest matches and whether two of three clinical geneticists would definitely have considered this diagnosis based on the image alone

Syndrome	Frequency	Correctly classified by neither	Correctly classified by both	Correctly classified by software, but not clinicians	Correctly classified by clinicians, but not software	McNemars Chi-square	p-value	Kappa
Williams	183	1	105	74	3	65.47	<.00001	−0.02
Rubinstein-Taybi	155	0	96	59	0	.	.	0.00
Floating Harbor	61	1	43	15	2	9.94	0.00162	0.02
Coffin Lowry	153	5	63	83	2	77.19	<.00001	0.02
Kabuki	195	1	152	40	2	34.38	<.00001	0.02
Smith Magenis	124	1	56	67	0	67.00	<.00001	0.01
PACS1	39	11	2	26	0	26.00	<.00001	0.04
Kleefstra	128	9	13	106	0	106.00	<.00001	0.02
Koolan-de Vries	120	8	58	51	3	42.67	<.00001	0.09
Cornelia de Lange	441	6	321	111	3	102.32	<.00001	0.06