Comparing Acoustic-based Approaches for Alzheimer's Disease Detection

2021/06/03 by Aparna Balagopalan, Balagopalan, Aparna, Jekaterina Novikova +1 · 4 citations
Computer Science · #Speech Recognition and Synthesis #Music and Audio Processing #Speech and Audio Processing

paper · pdf · doi:10.48550/arxiv.2106.01555

Abstract

Robust strategies for Alzheimer's disease (AD) detection are important, given the high prevalence of AD. In this paper, we study the performance and generalizability of three approaches for AD detection from speech on the recent ADReSSo challenge dataset: 1) using conventional acoustic features 2) using novel pre-trained acoustic embeddings 3) combining acoustic features and embeddings. We find that while feature-based approaches have a higher precision, classification approaches relying on pre-trained embeddings prove to have a higher, and more balanced cross-validated performance across multiple metrics of performance. Further, embedding-only approaches are more generalizable. Our best model outperforms the acoustic baseline in the challenge by 2.8%.

Cited by

• Foundation Model-based Evaluation of Neuropsychiatric Disorders: A Lifespan-Inclusive, Multi-Modal, and Multi-Lingual Study
• MoTAS: MoE-Guided Feature Selection from TTS-Augmented Speech for Enhanced Multimodal Alzheimer's Early Screening
• Recognizing Dementia from Neuropsychological Tests with State Space Models
• SoftStep: Learning Sparse Similarity Powers Deep Neighbor-Based Regression

Related

• Alzheimers Dementia Detection using Acoustic & Linguistic features and Pre-Trained BERT
• To BERT or Not To BERT: Comparing Speech and Language-based Approaches\n for Alzheimer's Disease Detection
• Exploring linguistic feature and model combination for speech recognition based automatic AD detection
• Leveraging Pretrained Representations with Task-related Keywords for Alzheimer's Disease Detection
• Beyond Manual Transcripts: The Potential of Automated Speech Recognition Errors in Improving Alzheimer's Disease Detection