Fig. 3 | Scientific Reports

Fig. 3

From: Alzheimer’s disease recognition using graph neural network by leveraging image-text similarity from vision language model

Fig. 3

Examples of bipartite graphs for AD and HC group samples, respectively. The darker edges represent higher edge weights. In the AD sample, the edges (image node, text node) = (2, 2), (6, 2), (10, 7), (2, 10), (6, 10) have higher weights, indicating a tendency for the text to focus on specific crop images. In contrast, in the HC sample, the edges (image node, text node) = (4, 1), (5, 3), (2, 4), (6, 4), (1, 5), (2, 5), (6, 5), (9, 5), (8, 7) have relatively higher weights, showing that the text is more evenly connected across multiple crop images.

Back to article page