A multimodal cross-transformer-based model to predict mild cognitive impairment using speech, language and vision

Farida Far Poor, Hiroko H. Dodge, Mohammad H. Mahoor

Abstract: Highlights•Novel cross-transformer model predicts MCI from speech, language, and vision data.•Multimodal integration improves early detection of mild cognitive impairment.•Cross-transformer architecture enhances interaction between speech, language, and vision.•Demonstrated superior accuracy in MCI prediction over unimodal and bimodal methods.