Identification of Cognitive Decline from Spoken Language through Feature Selection and the Bag of Acoustic Words Model

TL;DR

The paper tackles early detection of cognitive decline from naturally spoken language by focusing on acoustic paralinguistic cues rather than content. It combines feature selection on the eGeMAPS feature set with a bag-of-acoustic-words representation and a Chi2-SVM classifier to distinguish controls from dementia patients, validated on the Pitt Corpus and ADReSS 2020 data using LOSO cross-validation. A 25-feature subset—including MFCCs 1–4, F0 characteristics, and pausing metrics—achieves about 75% accuracy on the test set, demonstrating competitive performance with a compact, language-agnostic feature set. The work emphasizes reproducibility by addressing clustering randomness and demonstrates that non-lexical acoustic features can offer scalable, privacy-preserving diagnostic potential for cognitive impairment screening.

Abstract

Memory disorders are a central factor in the decline of functioning and daily activities in elderly individuals. The confirmation of the illness, initiation of medication to slow its progression, and the commencement of occupational therapy aimed at maintaining and rehabilitating cognitive abilities require a medical diagnosis. The early identification of symptoms of memory disorders, especially the decline in cognitive abilities, plays a significant role in ensuring the well-being of populations. Features related to speech production are known to connect with the speaker's cognitive ability and changes. The lack of standardized speech tests in clinical settings has led to a growing emphasis on developing automatic machine learning techniques for analyzing naturally spoken language. Non-lexical but acoustic properties of spoken language have proven useful when fast, cost-effective, and scalable solutions are needed for the rapid diagnosis of a disease. The work presents an approach related to feature selection, allowing for the automatic selection of the essential features required for diagnosis from the Geneva minimalistic acoustic parameter set and relative speech pauses, intended for automatic paralinguistic and clinical speech analysis. These features are refined into word histogram features, in which machine learning classifiers are trained to classify control subjects and dementia patients from the Dementia Bank's Pitt audio database. The results show that achieving a 75% average classification accuracy with only twenty-five features with the separate ADReSS 2020 competition test data and the Leave-One-Subject-Out cross-validation of the entire competition data is possible. The results rank at the top compared to international research, where the same dataset and only acoustic features have been used to diagnose patients.

Figures (3)

• Figure 1: Classification process over bag of acoustic word model.
• Figure 2: Average feature importances in sorted order for eGeMAPS feature set and relative speech pauses.
• Figure 3: LOSO cross-validation accuracy over 25 repetitions of replicated clustering.