Fig. 2: The scatterplots of predicted breast age and chronological age on Internal and external datasets.
From: Mammo-AGE: deep learning estimation of breast age from mammograms
The scatterplots demonstrate the high correlation relationship between the predicted breast age and chronological age, with density plots on the margins to show the distribution of predictions and actual ages. Pearson (r) correlation coefficients between predicted breast age and chronological age are also reported; MAE: mean absolute error; CS (α = 5 y): cumulative score based on a threshold of error within five years. The high Pearson correlation coefficients and relatively low MAE values across both internal and external datasets indicate the robustness and generalizability of the Mammo-AGE model in predicting breast age accurately.
