Skip to main content
[Preprint]. 2023 Sep 12:2023.09.12.23295132. [Version 1] doi: 10.1101/2023.09.12.23295132

Fig. 3. Prognostic performance of FaceAge in Several cancer cohorts.

Fig. 3.

a) Kaplan-Meier survival analysis shows significantly worse survival for increasing FaceAge estimation from only a face photograph as input. b) Forest plots of FaceAge risk groups demonstrate significant differences for all groups univariately, and this remains significant after adjustment for age, gender, and tumor site for the two oldest-looking groups compared to the youngest group as reference. c) Forest plots of FaceAge estimates as a continuous parameter for all patients as well as the four largest tumor sites. Note that FaceAge is significant in all tumor sites in univariate analysis and remains significant in breast, gastrointestinal, and genitourinary cancer after correction for age and gender. HR: Hazard Ratio; CI: confidence interval; (**: P < 0.01; ***: P < 0.001; all analyses are performed in the MAASTRO Cohort, n = 4,906)).