Mitigating Confounding in Speech-Based Dementia Detection through Weight Masking

TL;DR

This paper addresses gender confounding in speech-based dementia detection by proposing weight masking methods to isolate and remove gender-related biases in transformer models, improving fairness at some cost to accuracy.

Contribution

It introduces two novel weight masking techniques, the Extended Confounding Filter and Dual Filter, to mitigate gender bias in transformer-based dementia detection models.

Findings

Gender bias affects dementia detection models.

Weight masking reduces gender bias in model predictions.

Deconfounding slightly decreases detection accuracy.

Abstract

Deep transformer models have been used to detect linguistic anomalies in patient transcripts for early Alzheimer's disease (AD) screening. While pre-trained neural language models (LMs) fine-tuned on AD transcripts perform well, little research has explored the effects of the gender of the speakers represented by these transcripts. This work addresses gender confounding in dementia detection and proposes two methods: the $\textit{Extended Confounding Filter}$ and the $\textit{Dual Filter}$, which isolate and ablate weights associated with gender. We evaluate these methods on dementia datasets with first-person narratives from patients with cognitive impairment and healthy controls. Our results show transformer models tend to overfit to training data distributions. Disrupting gender-related weights results in a deconfounded dementia classifier, with the trade-off of slightly reduced dementia detection performance.