SpeechCARE: dynamic multimodal modeling for cognitive screening in diverse linguistic and speech task contexts - Takeaways - MDSpire

SpeechCARE: dynamic multimodal modeling for cognitive screening in diverse linguistic and speech task contexts

  • By

  • Hossein Azadmaleki

  • Yasaman Haghbin

  • Sina Rashidi

  • Mohammad Javad Momeni Nezhad

  • Ali Zolnour

  • Maryam Zolnoori

  • November 17, 2025

  • 0 min

Share

  • 1

    SpeechCARE is a multimodal transformer pipeline designed to detect cognitive impairment through brief speech recordings.

  • 2

    It achieved an average F1-score of 72.11% on the test set, demonstrating strong performance in classifying cognitive conditions.

  • 3

    The model integrates acoustic and linguistic features with demographic data using an Adaptive Gating Fusion mechanism.

  • 4

    SpeechCARE complements traditional biomarkers by capturing functional speech deficits for early detection of cognitive decline.

  • 5

    The model shows strong multilingual generalizability, although fairness analysis indicated moderate disparities for Spanish speakers.

Original Source(s)

SpeechCARE: dynamic multimodal modeling for cognitive screening in diverse linguistic and speech task contexts

  • by Hossein Azadmaleki, Yasaman Haghbin, Sina Rashidi, Mohammad Javad Momeni Nezhad, Ali Zolnour, Maryam Zolnoori

  • November 17, 2025

Related Content

Patterns of Initial Antidementia Medication Prescriptions Among Community-Dwelling Individuals in Norway: A Comprehensive Register-Based Analysis

Targeting the miR-32533/CREB5 axis: a promising immunomodulatory and therapeutic strategy for Alzheimer’s disease

Lead Exposure Implicated in Rising Alzheimer's Rates

US study finds that patients with high bone lead levels are three times more likely to develop the disease

  • February 19, 2026

  • 2 min