Deep-learning analysis of speech using mel-spectrograms for the assessment of mild cognitive impairment and Alzheimer's disease

BackgroundSpeech abnormalities are recognized as early indicators of Alzheimer's disease (AD) and mild cognitive impairment (MCI).ObjectiveTo determine whether deep-learning models trained on mel-spectrograms of brief speech tasks can (i) discriminate individuals with MCI and AD from cognitively normal controls (NC) and (ii) estimate cognitive status with clinically useful accuracy.MethodsSpeech from 594 participants (185 NC, 231 MCI, 178 AD) was recorded through a mobile application that included 11 cognitive-linguistic tasks. Audio was converted into mel-spectrogram images and processed using a VGG16-based deep-learning model with transfer learning and fine-tuning of block 5. Task-specific feature vectors were extracted, concatenated, and used to train a deep neural network. The dataset was split into training, validation, and test sets (3:1:1), and five-split cross-validation was performed.ResultsThe model demonstrated an overall accuracy of 72.4% in classifying NC from the abnormal group (MCI and AD), with sensitivity and specificity of 72.5% and 72.2%, respectively, a balanced accuracy of 72.4%, and an AUC of 0.997. In binary classifications, the model achieved 82.9% accuracy (balanced accuracy 82.9%, AUC 0.992) for NC versus AD, 70.7% accuracy (balanced accuracy 70.3%, AUC 0.956) for NC versus MCI, and 77.5% accuracy (balanced accuracy 78.9%, AUC 0.889) for MCI versus AD. Tasks such as serial subtraction, storytelling, and picture description contributed most to classification performance, indicating their effectiveness in capturing cognitive deficits.ConclusionsMel-spectrogram-based deep-learning analysis of speech shows promise as a rapid, non-invasive, and language-independent screening tool for early cognitive impairment, with potential advantages over traditional assessments such as the Mini-Mental State Examination.

Authors

Institutions

Publication Details

Journal
Journal of Alzheimer s Disease
Published
2025-12-12
DOI
https://doi.org/10.1177/13872877251401202
Primary Topic
Dementia and Cognitive Impairment Research
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Deep-learning analysis of speech using mel-spectrograms for the assessment of mild cognitive impairment and Alzheimer's disease

Seunghee Na, Yun Ho Choi, Seon Young Ryu, YongSoo Shim et al.
Journal of Alzheimer s Disease
Dementia and Cognitive Impairment Research
article

Deep-learning analysis of speech using mel-spectrograms for the assessment of mild cognitive impairment and Alzheimer's disease

Seunghee Na, Yun Ho Choi, Seon Young Ryu, YongSoo Shim, A-Hyun Cho, Hojin Choi, Hyun Kook Lim, Woojun Kim, Eun Ye Lim, Dong Yang, Hae‐Eun Shin, Bora Yoon, Seonjeong Byun, Bohyun Wang, S. Lee Hong, In‐Uk Song, Yun Jeong Hong, S.S. Park, Caroline Baek, Changtae Hahn, Han Jo Kim, Hyung-Jun Kim, Dong Woo Kang, Joon S. Lim, Wang-Yeon Won
article en

Abstract

BackgroundSpeech abnormalities are recognized as early indicators of Alzheimer's disease (AD) and mild cognitive impairment (MCI).ObjectiveTo determine whether deep-learning models trained on mel-spectrograms of brief speech tasks can (i) discriminate individuals with MCI and AD from cognitively normal controls (NC) and (ii) estimate cognitive status with clinically useful accuracy.MethodsSpeech from 594 participants (185 NC, 231 MCI, 178 AD) was recorded through a mobile application that included 11 cognitive-linguistic tasks. Audio was converted into mel-spectrogram images and processed using a VGG16-based deep-learning model with transfer learning and fine-tuning of block 5. Task-specific feature vectors were extracted, concatenated, and used to train a deep neural network. The dataset was split into training, validation, and test sets (3:1:1), and five-split cross-validation was performed.ResultsThe model demonstrated an overall accuracy of 72.4% in classifying NC from the abnormal group (MCI and AD), with sensitivity and specificity of 72.5% and 72.2%, respectively, a balanced accuracy of 72.4%, and an AUC of 0.997. In binary classifications, the model achieved 82.9% accuracy (balanced accuracy 82.9%, AUC 0.992) for NC versus AD, 70.7% accuracy (balanced accuracy 70.3%, AUC 0.956) for NC versus MCI, and 77.5% accuracy (balanced accuracy 78.9%, AUC 0.889) for MCI versus AD. Tasks such as serial subtraction, storytelling, and picture description contributed most to classification performance, indicating their effectiveness in capturing cognitive deficits.ConclusionsMel-spectrogram-based deep-learning analysis of speech shows promise as a rapid, non-invasive, and language-independent screening tool for early cognitive impairment, with potential advantages over traditional assessments such as the Mini-Mental State Examination.

Journal of Alzheimer s DiseaseVol. 109(2)
Gachon University (KR), St. Mary's Hospital (US), The Catholic University of Korea St. Vincent's Hospital (KR), The Catholic University of Korea Incheon St. Mary's Hospital (KR), Konyang University Hospital (KR), The Catholic University of Korea Uijeongbu St. Mary's Hospital (KR), The Catholic University of Korea Yeouido St. Mary's Hospital (KR), St. Mary's Hospital (US), Hanyang University Guri Hospital (KR), St Mary's Hospital (GB), The Catholic University of Korea Bucheon St. Mary's Hospital (KR), Hanyang University Seoul Hospital (KR), Io Therapeutics (United States) (US), The Catholic University of Korea Seoul St. Mary's Hospital (KR), Catholic University of Korea (KR)
Openalex Percentile: Top 53%
Dementia and Cognitive Impairment Research
DOISourceOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.