highly significantp < 0.001
When the cross-modal attention mechanism was introduced on top of the S-I module, it showed highly significant (p < 0.001, between CNN and CNN + S-I-C module) accuracy, with very significant (p < 0.01) AUC and F-1 scores.
When the cross-modal attention mechanism was introduced on top of the S-I module, it showed highly significant (p < 0.001, between CNN and CNN + S-I-C module) accuracy, with very significant (p < 0.01) AUC and F-1 scores.