MedVoiceBias: A Controlled Study of Audio LLM Behavior in Clinical Decision-Making

• URL: http://arxiv.org/abs/2511.06592v1
• Date: Mon, 10 Nov 2025 00:44:37 GMT
• Title: MedVoiceBias: A Controlled Study of Audio LLM Behavior in Clinical Decision-Making
• Abstract summary: Large language models transition from text-based interfaces to audio interactions in clinical settings.<n>We evaluated these models on 170 clinical cases, each synthesized into speech from 36 distinct voice profiles.
• Score: 24.535496581050094
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: As large language models transition from text-based interfaces to audio interactions in clinical settings, they might introduce new vulnerabilities through paralinguistic cues in audio. We evaluated these models on 170 clinical cases, each synthesized into speech from 36 distinct voice profiles spanning variations in age, gender, and emotion. Our findings reveal a severe modality bias: surgical recommendations for audio inputs varied by as much as 35% compared to identical text-based inputs, with one model providing 80% fewer recommendations. Further analysis uncovered age disparities of up to 12% between young and elderly voices, which persisted in most models despite chain-of-thought prompting. While explicit reasoning successfully eliminated gender bias, the impact of emotion was not detected due to poor recognition performance. These results demonstrate that audio LLMs are susceptible to making clinical decisions based on a patient's voice characteristics rather than medical evidence, a flaw that risks perpetuating healthcare disparities. We conclude that bias-aware architectures are essential and urgently needed before the clinical deployment of these models.

Related papers

• Systematic Study of Dysarthric Speech Recognition: Spectral Features and Acoustic Models [48.10903631277461]
  This paper presents a comprehensive investigation of various combinations of acoustic features tailored to different Acoustic Models.<n>The incorporation of Pitch features notably improved recognition performance, especially for sentence recognition tasks.<n>We have demonstrated the potential to enhance the performance of the state-of-the-art Factorized Time Delay Neural Network (F-TDNN) model for recognizing dysarthric speech.
  arXiv  Detail & Related papers  (2026-06-18T04:57:37Z)
• A Proactive Multi-Agent Dialogue Framework for Assessing Social Language Disorder Traits in Autism [11.474807247446526]
  We present TPA (Think, Plan, Ask), a proactive multi-agent dialogue framework applied to the language assessment component of the Autism Diagnostic Observation Schedule Module 4 (ADOS-2)<n>A doctor agent explicitly reasons about which traits remain unobserved before selecting a clinically grounded strategy and generating a targeted question.<n>A patient agent grounded in real ADOS-2 clinical data enables reproducible evaluation without real patient participation, validated across three independent experiments confirming adequate fidelity to real patient language.
  arXiv  Detail & Related papers  (2026-05-21T19:45:46Z)
• MedArena: Comparing LLMs for Medicine-in-the-Wild Clinician Preferences [50.71326426975699]
  We present MedArena, an interactive evaluation platform for medical large language models (LLMs)<n>Given a clinician-provided query, MedArena presents responses from two randomly selected models and asks the user to select the preferred response.<n>Out of 1571 preferences collected across 12 LLMs up to November 1, 2025, Gemini 2.0 Flash Thinking, Gemini 2.5 Pro, and GPT-4o were the top three models by Bradley-Terry rating.
  arXiv  Detail & Related papers  (2026-03-13T22:30:26Z)
• CO-VADA: A Confidence-Oriented Voice Augmentation Debiasing Approach for Fair Speech Emotion Recognition [49.27067541740956]
  We present CO-VADA, a Confidence-Oriented Voice Augmentation Debiasing Approach that mitigates bias without modifying model architecture or relying on demographic information.<n>CO-VADA identifies training samples that reflect bias patterns present in the training data and then applies voice conversion to alter irrelevant attributes and generate samples.<n>Our framework is compatible with various SER models and voice conversion tools, making it a scalable and practical solution for improving fairness in SER systems.
  arXiv  Detail & Related papers  (2025-06-06T13:25:56Z)
• A Multi-modal Approach to Dysarthria Detection and Severity Assessment Using Speech and Text Information [9.172160338245252]
  This study introduces a novel approach that leverages both speech and text modalities.<n>By employing cross-attention mechanism, our method learns the acoustic and linguistic similarities between speech and text representations.
  arXiv  Detail & Related papers  (2024-12-22T06:08:35Z)
• A Classification Benchmark for Artificial Intelligence Detection of Laryngeal Cancer from Patient Voice [0.30723404270319693]
  Current diagnostic pathways are inefficient, putting undue stress on both patients and the medical system.<n>We introduce a benchmark suite comprising 36 models trained and evaluated on open-source datasets.<n>Our best model achieves a balanced accuracy of 83.7%, sensitivity of 84.0%, specificity of 83.3%, and AUROC of 91.8%.
  arXiv  Detail & Related papers  (2024-12-20T10:34:03Z)
• Cognitive Insights Across Languages: Enhancing Multimodal Interview Analysis [0.6062751776009752]
  We propose a multimodal model capable of predicting Mild Cognitive Impairment and cognitive scores. The proposed model demonstrates the ability to transcribe and differentiate between languages used in the interviews. Our approach involves in-depth research to implement various features obtained from the proposed modalities.
  arXiv  Detail & Related papers  (2024-06-11T17:59:31Z)
• Show from Tell: Audio-Visual Modelling in Clinical Settings [58.88175583465277]
  We consider audio-visual modelling in a clinical setting, providing a solution to learn medical representations without human expert annotation. A simple yet effective multi-modal self-supervised learning framework is proposed for this purpose. The proposed approach is able to localise anatomical regions of interest during ultrasound imaging, with only speech audio as a reference.
  arXiv  Detail & Related papers  (2023-10-25T08:55:48Z)
• A Transformer-based representation-learning model with unified processing of multimodal input for clinical diagnostics [63.106382317917344]
  We report a Transformer-based representation-learning model as a clinical diagnostic aid that processes multimodal input in a unified manner. The unified model outperformed an image-only model and non-unified multimodal diagnosis models in the identification of pulmonary diseases.
  arXiv  Detail & Related papers  (2023-06-01T16:23:47Z)
• Exploiting prompt learning with pre-trained language models for Alzheimer's Disease detection [70.86672569101536]
  Early diagnosis of Alzheimer's disease (AD) is crucial in facilitating preventive care and to delay further progression. This paper investigates the use of prompt-based fine-tuning of PLMs that consistently uses AD classification errors as the training objective function.
  arXiv  Detail & Related papers  (2022-10-29T09:18:41Z)
• Exploring linguistic feature and model combination for speech recognition based automatic AD detection [61.91708957996086]
  Speech based automatic AD screening systems provide a non-intrusive and more scalable alternative to other clinical screening techniques. Scarcity of specialist data leads to uncertainty in both model selection and feature learning when developing such systems. This paper investigates the use of feature and model combination approaches to improve the robustness of domain fine-tuning of BERT and Roberta pre-trained text encoders.
  arXiv  Detail & Related papers  (2022-06-28T05:09:01Z)
• Continuous Speech for Improved Learning Pathological Voice Disorders [12.867900671251395]
  This study proposes a novel approach, using continuous Mandarin speech instead of a single vowel, to classify four common voice disorders. In the proposed framework, acoustic signals are transformed into mel-frequency cepstral coefficients, and a bi-directional long-short term memory network (BiLSTM) is adopted to model the sequential features.
  arXiv  Detail & Related papers  (2022-02-22T09:58:31Z)
• What Do You See in this Patient? Behavioral Testing of Clinical NLP Models [69.09570726777817]
  We introduce an extendable testing framework that evaluates the behavior of clinical outcome models regarding changes of the input. We show that model behavior varies drastically even when fine-tuned on the same data and that allegedly best-performing models have not always learned the most medically plausible patterns.
  arXiv  Detail & Related papers  (2021-11-30T15:52:04Z)
• Comparison of Speaker Role Recognition and Speaker Enrollment Protocol for conversational Clinical Interviews [9.728371067160941]
  We train end-to-end neural network architectures to adapt to each task and evaluate each approach under the same metric. Results do not depend on the demographics of the Interviewee, highlighting the clinical relevance of our methods.
  arXiv  Detail & Related papers  (2020-10-30T09:07:37Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.