Skip to main content

View full-text article in PMC

. 2022 Jun 21;5(6):e2217854. doi: 10.1001/jamanetworkopen.2022.17854

Table. Comparisons of Model Screening Performance in Training and Validation Sets.

Model performance	Measure (95% CI)
Model performance	AUC	Accuracy	Sensitivity	Specificity
NT^a
Training	0.78 (0.73-0.83)	0.84 (0.80-0.87)	0.94 (0.91-0.96)	0.62 (0.55-0.70)
Validation	0.69 (0.61-0.76)	0.79 (0.73-0.83)	0.47 (0.37-0.58)	0.95 (0.91-0.98)
NT + age^b
Training	0.82 (0.77-0.86)	0.82 (0.78-0.85)	0.67 (0.59-0.74)	0.89 (0.85-0.92)
Validation	0.73 (0.66-0.80)	0.76 (0.71-0.81)	0.52 (0.41-0.62)	0.89 (0.83-0.93)
DL
Training	0.98 (0.97-0.99)	0.94 (0.92-0.96)	0.95 (0.91-0.98)	0.93 (0.90-0.96)
Validation	0.95 (0.93-0.98)	0.88 (0.84-0.92)	0.76 (0.66-0.85)	0.94 (0.90-0.97)
DL + age^c
Training	0.98 (0.97-0.99)	0.94 (0.91-0.96)	0.95 (0.91-0.98)	0.93 (0.90-0.95)
Validation	0.95 (0.93-0.98)	0.89 (0.84-0.92)	0.78 (0.69-0.86)	0.94 (0.89-0.97)

Abbreviations: AUC, area under the receiver operating characteristic curve; DL, deep learning; NT, nuchal translucency.

^{^a}

Model constructed based on fetal NT.

^{^b}

Model constructed based on fetal NT and maternal age.

^{^c}

Model constructed by DL integrating maternal age.