# The Unreliability of Acoustic Systems in Alzheimer's Speech Datasets with Heterogeneous Recording Conditions

arXiv: https://arxiv.org/abs/2409.12170 · HTML: https://mlsanity.com/paper/2409.12170/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 38 references that are in the map.
- First posted on arXiv 11 Sep 2024.

## Builds on (most cited first, 8 of 8)

- [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://mlsanity.com/paper/2006.11477/) (2020)
- [Emotion Recognition from Speech Using Wav2vec 2.0 Embeddings](https://mlsanity.com/paper/2104.03502/) (2021)
- [Towards Learning a Universal Non-Semantic Representation of Speech](https://mlsanity.com/paper/2002.12764/) (2020)
- [Detecting cognitive decline using speech only: The ADReSSo Challenge](https://mlsanity.com/paper/2104.09356/) (2021)
- [Speaker Recognition Based on Deep Learning: An Overview](https://mlsanity.com/paper/2012.00931/) (2020)
- [On the use of Self-supervised Pre-trained Acoustic and Linguistic Features for Continuous Speech Emotion Recognition](https://mlsanity.com/paper/2011.09212/) (2020)
- [Comparing Natural Language Processing Techniques for Alzheimer's Dementia Prediction in Spontaneous Speech](https://mlsanity.com/paper/2006.07358/) (2020)
- [Recognizing More Emotions with Less Data Using Self-supervised Transfer Learning](https://mlsanity.com/paper/2011.05585/) (2020)

## Built on by (most cited first, 4 of 4)

- [Shortcut Learning in Binary Classifier Black Boxes: Applications to Voice Anti-Spoofing and Biometrics](https://mlsanity.com/paper/2601.17782/) (2026)
- [Good practices for evaluation of machine learning systems](https://mlsanity.com/paper/2412.03700/) (2024)
- [Transcript-Free Lightweight Detection of Alzheimer's Disease from Spontaneous Speech Using Handcrafted MFCC-Dominant Acoustic Biomarkers](https://mlsanity.com/paper/2607.10168/) (2026)
- [Beyond Decodability: Do Acoustic Factors Drive Predictions in Speech-Based Alzheimer's Assessment?](https://mlsanity.com/paper/2610.01846/) (2026)

## Abstract (arXiv)

Automated speech analysis is a thriving approach to detect early markers of Alzheimer's disease (AD). Yet, recording conditions in most AD datasets are heterogeneous, with patients and controls often evaluated in different acoustic settings. While this is not a problem for analyses based on speech transcription or features obtained from manual alignment, it does cast serious doubts on the validity of acoustic features, which are strongly influenced by acquisition conditions. We examined this issue in the ADreSSo dataset, derived from the widely used Pitt corpus. We show that systems based on two acoustic features, MFCCs and Wav2vec 2.0 embeddings, can discriminate AD patients from controls with above-chance performance when using only the non-speech part of the audio signals. We replicated this finding in a separate dataset of Spanish speakers. Thus, in these datasets, the class can be partly predicted by recording conditions. Our results are a warning against the use of acoustic systems for identifying patients based on non-standardized recordings. We propose that acoustically heterogeneous datasets for dementia studies should be either (a) analyzed using only transcripts or other features derived from manual annotations, or (b) replaced by datasets collected with strictly controlled acoustic conditions.