Alzheimer's Dementia Recognition Using Acoustic, Lexical, Disfluency and Speech Pause Features Robust to Noisy Inputs

TL;DR

This paper introduces multimodal deep learning models that combine acoustic, lexical, disfluency, and pause features from speech to accurately detect Alzheimer's Disease and assess cognitive decline, demonstrating robustness to noisy inputs.

Contribution

The study presents a novel multimodal fusion approach using deep learning that improves Alzheimer's detection accuracy and cognitive score prediction, especially in noisy conditions.

Findings

Achieved 84% accuracy in AD classification.

Reduced RMSE to 4.26 in cognitive score prediction.

Demonstrated robustness to noisy ASR inputs.

Abstract

We present two multimodal fusion-based deep learning models that consume ASR transcribed speech and acoustic data simultaneously to classify whether a speaker in a structured diagnostic task has Alzheimer's Disease and to what degree, evaluating the ADReSSo challenge 2021 data. Our best model, a BiLSTM with highway layers using words, word probabilities, disfluency features, pause information, and a variety of acoustic features, achieves an accuracy of 84% and RSME error prediction of 4.26 on MMSE cognitive scores. While predicting cognitive decline is more challenging, our models show improvement using the multimodal approach and word probabilities, disfluency and pause information over word-only models. We show considerable gains for AD classification using multimodal fusion and gating, which can effectively deal with noisy inputs from acoustic features and ASR hypotheses.