Skip to main content
Fig. 4 | Journal of Translational Medicine

Fig. 4

From: Predicting long-term progression of Alzheimer’s disease using a multimodal deep learning model incorporating interaction effects

Fig. 4

Comparison of model performance with and without interaction effects and multimodality. A Effectiveness evaluation of the dual interaction modules. The performance of our DISFC model was compared with the simple fusion benchmark model without intra-modal and inter-modal interaction modules. The box plot illustrates the 25th percentile (upper box limit), median (horizontal centerline), and 75th percentile (lower box limit). The upper whisker, lower whisker, and hollow circle symbol indicate the maximum, minimum, and mean values of a given model for each metric, respectively. The shaded area on one side around each box represents the probability density. B Performance comparison of the models based on different modality combinations. The unimodal, bimodal, and trimodal models were cross-validated with identical settings. Each bar represents the mean value across folds for each metric

Back to article page