PROCESS-2: A Benchmark Speech Corpus for Early Cognitive Impairment Detection

2026-05-14

AI Summary

This study addresses the scarcity of large-scale, clinically validated speech datasets collected in real-world settings by introducing PROCESS-2, a novel resource built upon the standardized digital cognitive assessment platform CognoMemory. The dataset comprises 21 hours of spontaneous speech from 400 participants performing picture description and verbal fluency tasks, accompanied by manual transcriptions, clinical diagnostic labels, and rich metadata, with predefined train/test splits. PROCESS-2 is the first to simultaneously ensure conversational diversity and clinical reliability in authentic environments, while employing a controlled-access framework to balance open sharing with privacy preservation. Baseline experiments demonstrate that models trained on this dataset effectively differentiate between cognitively healthy individuals, those with mild cognitive impairment, and dementia patients, thereby validating PROCESS-2 as a robust and reproducible benchmark for speech-based cognitive impairment assessment.

Abstract

Speech-based analysis offers a scalable and non-invasive approach for detecting cognitive decline, yet progress has been constrained by the limited availability of clinically validated datasets collected under realistic conditions. We introduce PROCESS-2, a large-scale speech dataset designed to support research on automatic assessment of cognitive impairment from spontaneous and task-oriented speech. The dataset comprises recordings from 200 healthy controls, 150 mild cognitive impairment, and 50 dementia diagnoses collected using the CognoMemory digital assessment platform. Each participant completed a single assessment session, including picture description and verbal fluency tasks, accompanied by manually verified transcripts and participant-level metadata. PROCESS-2 contains approximately 21 hours of speech audio with predefined train/test partitions. Comprehensive technical validation evaluated demographic balance, clinical consistency, recording stability, embedding-space structure, and reproducible baseline modelling performance, demonstrating clinically meaningful group separation and stable performance across modelling approaches while preserving real-world conversational variability. PROCESS-2 is released under controlled access via Hugging Face to enable responsible reuse while protecting participant privacy, providing a reproducible benchmark resource for speech-based cognitive assessment research.

Problem

• cognitive impairment detection
• speech corpus
• clinically validated dataset
• realistic conditions
• benchmark

Innovation

• speech-based cognitive assessment
• large-scale speech corpus
• mild cognitive impairment detection
• clinically validated dataset
• reproducible benchmark