{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,21]],"date-time":"2026-05-21T18:12:26Z","timestamp":1779387146191,"version":"3.53.1"},"reference-count":37,"publisher":"Public Library of Science (PLoS)","issue":"5","license":[{"start":{"date-parts":[[2026,5,21]],"date-time":"2026-05-21T00:00:00Z","timestamp":1779321600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100013129","name":"Ministry of SMEs and Startups","doi-asserted-by":"publisher","award":["S3079103"],"award-info":[{"award-number":["S3079103"]}],"id":[{"id":"10.13039\/501100013129","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Agriculture Science and Technology Development Rural Development Administration","award":["PJ01712403"],"award-info":[{"award-number":["PJ01712403"]}]}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:sec id=\"sec001\">\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>The early detection of cognitive impairments, such as mild cognitive impairment (MCI) and Alzheimer\u2019s disease (AD), is essential for timely intervention and management. This study evaluates the performance of various deep-learning models in classifying speech recordings from individuals with normal cognition (NC), MCI, and AD, to identify the most effective approach for audio-based cognitive impairment diagnosis.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec002\">\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>Speech data were obtained from the AI Hub \u201cCognitive Impairment Diagnosis Voice\/Conversation\u201d dataset. The study analyzed voice recordings from 320 female participants (105 with Alzheimer\u2019s disease, 92 with mild cognitive impairment, and 123 cognitively normal controls). Three deep-learning architectures were compared: a one-dimensional convolutional neural network (1D CNN), an audio spectrogram transformer (AST), and a speech recognition model (Wav2Vec 2.0). The models were trained using features such as spectrograms, mel-spectrograms, and mel-frequency cepstral coefficients (MFCCs). Model performance was assessed using accuracy, recall, precision, and F1-score, with a five-fold cross-validation strategy to ensure robust and unbiased evaluation. Statistical significance was assessed using pairwise proportion z-tests with Holm-Bonferroni correction, and Wilson score 95% confidence intervals were computed for each model\u2019s accuracy.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec003\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Wav2Vec 2.0 outperformed the other models, achieving the highest accuracy and F1-scores for NC vs. MCI (accuracy: 0.74, F1: 0.72) and NC vs. AD (accuracy: 0.83, F1: 0.83). Pairwise proportion z-tests with Holm-Bonferroni correction confirmed that Wav2Vec 2.0 significantly outperformed 7 of 10 competing models (corrected p\u2009&lt;\u20090.05) in both classification tasks. Performance varied by model and classification task, with Wav2Vec 2.0 consistently demonstrating superior accuracy across labels.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec004\">\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>This study emphasizes the importance of selecting appropriate models and features for task-specific optimization and provides a foundation for developing non-invasive, speech-based diagnostic tools for cognitive disorders.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1371\/journal.pone.0349412","type":"journal-article","created":{"date-parts":[[2026,5,21]],"date-time":"2026-05-21T17:40:05Z","timestamp":1779385205000},"page":"e0349412","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Prediction of cognitive impairment through speech data analysis: A comparative evaluation of deep learning models"],"prefix":"10.1371","volume":"21","author":[{"given":"Minsoo","family":"Kim","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hyunjoo","family":"Choi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"YongSoo","family":"Shim","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nayoung","family":"Ryoo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ho Tae","family":"Jeong","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Gihyun","family":"Yun","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hunboc","family":"Lee","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9101-5704","authenticated-orcid":true,"given":"SangYun","family":"Kim","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Young Chul","family":"Youn","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,5,21]]},"reference":[{"key":"pone.0349412.ref001","doi-asserted-by":"crossref","first-page":"39","DOI":"10.1186\/s13195-016-0207-9","article-title":"Drug development in Alzheimer\u2019s disease: The path to 2025","volume":"8","author":"J Cummings","year":"2016","journal-title":"Alzheimers Res Ther"},{"issue":"4","key":"pone.0349412.ref002","doi-asserted-by":"crossref","first-page":"107","DOI":"10.12779\/dnd.2022.21.4.107","article-title":"Need for an update for the guideline for the management of mild cognitive impairment","volume":"21","author":"M Kim","year":"2022","journal-title":"Dement Neurocogn Disord"},{"key":"pone.0349412.ref003","article-title":"Mild Cognitive Impairment","volume-title":"StatPearls","author":"S Anand","year":"2024"},{"issue":"2","key":"pone.0349412.ref004","doi-asserted-by":"crossref","first-page":"518","DOI":"10.3390\/jcm13020518","article-title":"Mild cognitive impairment progression and Alzheimer\u2019s disease risk: A comprehensive analysis of 3553 cases over 203 months","volume":"13","author":"N \u00d6ks\u00fcz","year":"2024","journal-title":"J Clin Med"},{"issue":"2","key":"pone.0349412.ref005","doi-asserted-by":"crossref","first-page":"232","DOI":"10.1111\/jgs.15642","article-title":"Mild cognitive impairment that does not progress to dementia: A population-based study","volume":"67","author":"M Ganguli","year":"2019","journal-title":"J Am Geriatr Soc"},{"issue":"1","key":"pone.0349412.ref006","doi-asserted-by":"crossref","first-page":"57","DOI":"10.3233\/JAD-150786","article-title":"Reversible mild cognitive impairment: The role of comorbidities at baseline evaluation","volume":"51","author":"G Grande","year":"2016","journal-title":"J Alzheimers Dis"},{"key":"pone.0349412.ref007","article-title":"Review on pathogenesis and treatment of Alzheimer\u2019s disease","author":"J Cai","year":"2024","journal-title":"Dev Dyn"},{"issue":"4","key":"pone.0349412.ref008","doi-asserted-by":"crossref","first-page":"627","DOI":"10.3233\/JAD-2011-110458","article-title":"Review and meta-analysis of biomarkers and diagnostic imaging in Alzheimer\u2019s disease","volume":"26","author":"LM Bloudek","year":"2011","journal-title":"J Alzheimers Dis"},{"issue":"1","key":"pone.0349412.ref009","doi-asserted-by":"crossref","first-page":"16","DOI":"10.12779\/dnd.2023.22.1.16","article-title":"Diagnostic tools for Alzheimer\u2019s disease: A narrative review based on our own research experience","volume":"22","author":"SH Park","year":"2023","journal-title":"Dement Neurocogn Disord"},{"issue":"8","key":"pone.0349412.ref010","doi-asserted-by":"crossref","first-page":"65","DOI":"10.1186\/s13073-014-0065-7","article-title":"Blood biomarkers for Alzheimer\u2019s disease","volume":"6","author":"S Lovestone","year":"2014","journal-title":"Genome Med"},{"issue":"3","key":"pone.0349412.ref011","doi-asserted-by":"crossref","first-page":"158","DOI":"10.14283\/jpad.2020.19","article-title":"Rationale for early diagnosis of mild cognitive impairment (MCI) supported by emerging digital technologies","volume":"7","author":"MN Sabbagh","year":"2020","journal-title":"J Prev Alzheimers Dis"},{"key":"pone.0349412.ref012","article-title":"Evaluation of a speech-based AI system for early detection of Alzheimer\u2019s disease remotely via smartphones","author":"E Fristed","year":"2021","journal-title":"medRxiv"},{"issue":"1","key":"pone.0349412.ref013","doi-asserted-by":"crossref","first-page":"186","DOI":"10.1186\/s13195-022-01131-3","article-title":"Deep learning-based speech analysis for Alzheimer\u2019s disease detection: A literature review","volume":"14","author":"Q Yang","year":"2022","journal-title":"Alzheimers Res Ther"},{"key":"pone.0349412.ref014","author":"Y Gong","year":"2021"},{"key":"pone.0349412.ref015","first-page":"12449","article-title":"wav2vec 2.0: A framework for self-supervised learning of speech representations","volume":"33","author":"A Baevski","year":"2020","journal-title":"Advances in Neural Information Processing Systems"},{"key":"pone.0349412.ref016","author":"O Colliot","year":"2023","journal-title":"Machine Learning for Brain Disorders"},{"issue":"1","key":"pone.0349412.ref017","doi-asserted-by":"crossref","first-page":"15","DOI":"10.1016\/j.jvoice.2020.02.009","article-title":"Classification of voice disorders using a one-dimensional convolutional neural network","volume":"36","author":"S Fujimura","year":"2022","journal-title":"J Voice"},{"key":"pone.0349412.ref018","doi-asserted-by":"crossref","first-page":"1178","DOI":"10.1109\/OJCS.2025.3580570","article-title":"Radio frequency sensing-based human emotion identification by leveraging 2D transformation techniques and deep learning models","volume":"6","author":"N AbuAli","year":"2025","journal-title":"IEEE Open Journal of the Computer Society"},{"issue":"16","key":"pone.0349412.ref019","doi-asserted-by":"crossref","first-page":"5381","DOI":"10.3390\/s21165381","article-title":"Classification and visualisation of normal and abnormal radiographs; A comparison between eleven convolutional neural network architectures","volume":"21","author":"A Ananda","year":"2021","journal-title":"Sensors (Basel)"},{"key":"pone.0349412.ref020","unstructured":"CNN Architectures: LeNet, AlexNet, VGG, GoogLeNet, ResNet and More. https:\/\/coderzpy.com\/cnn-architectures-lenet-alexnet-vgg-googlenet-resnet-and-more\/"},{"key":"pone.0349412.ref021","author":"S Latif","year":"2023"},{"issue":"1","key":"pone.0349412.ref022","doi-asserted-by":"crossref","first-page":"87","DOI":"10.1109\/TPAMI.2022.3152247","article-title":"A survey on vision transformer","volume":"45","author":"K Han","year":"2023","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"pone.0349412.ref023","author":"T Mikolov","year":"2013"},{"key":"pone.0349412.ref024","author":"L Wyse","year":"2017"},{"issue":"5","key":"pone.0349412.ref025","doi-asserted-by":"crossref","first-page":"1231","DOI":"10.1109\/TSP.2003.810293","article-title":"Short-time fourier transform: two fundamental properties and an optimal implementation","volume":"51","author":"L Durak","year":"2003","journal-title":"IEEE Trans Signal Process"},{"issue":"18","key":"pone.0349412.ref026","doi-asserted-by":"crossref","first-page":"1","DOI":"10.5120\/16891-6927","article-title":"Comparative performance analysis of hamming, hanning and blackman window","volume":"96","author":"P Podder","year":"2014","journal-title":"IJCA"},{"issue":"6","key":"pone.0349412.ref027","doi-asserted-by":"crossref","first-page":"1368","DOI":"10.1137\/0914081","article-title":"Fast fourier transforms for nonequispaced data","volume":"14","author":"A Dutt","year":"1993","journal-title":"SIAM J Sci Comput"},{"key":"pone.0349412.ref028","doi-asserted-by":"crossref","first-page":"4779","DOI":"10.1109\/ICASSP.2018.8461368","article-title":"Natural TTS Synthesis by Conditioning Wavenet on MEL Spectrogram Predictions","author":"J Shen","year":"2018","journal-title":"2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)"},{"key":"pone.0349412.ref029","author":"L Muda","year":"2010"},{"key":"pone.0349412.ref030","author":"A Conneau","year":"2020"},{"key":"pone.0349412.ref031","volume-title":"ImageNet Classification with Deep Convolutional Neural Networks","author":"A Krizhevsky","year":"2012"},{"key":"pone.0349412.ref032","doi-asserted-by":"crossref","first-page":"2261","DOI":"10.1109\/CVPR.2017.243","article-title":"Densely connected convolutional networks","author":"G Huang","year":"2017","journal-title":"2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)"},{"key":"pone.0349412.ref033","doi-asserted-by":"crossref","first-page":"770","DOI":"10.1109\/CVPR.2016.90","article-title":"Deep residual learning for image recognition","author":"K He","year":"2016","journal-title":"2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)"},{"key":"pone.0349412.ref034","unstructured":"Simonyan K, Andrew Z. Very Deep Convolutional Networks for Large-Scale Image Recognition. 2014."},{"key":"pone.0349412.ref035","author":"H Ji","year":"2022"},{"key":"pone.0349412.ref036","article-title":"The optimization of a natural language processing approach for the automatic detection of Alzheimer\u2019s disease using GPT embeddings","author":"BS Runde","year":"2024","journal-title":"medRxiv"},{"issue":"12","key":"pone.0349412.ref037","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pdig.0000168","article-title":"Predicting dementia from spontaneous speech using large language models","volume":"1","author":"F Agbavor","year":"2022","journal-title":"PLOS Digit Health"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0349412","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,21]],"date-time":"2026-05-21T17:40:12Z","timestamp":1779385212000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0349412"}},"subtitle":[],"editor":[{"given":"Farman","family":"Ullah","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,5,21]]},"references-count":37,"journal-issue":{"issue":"5","published-online":{"date-parts":[[2026,5,21]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0349412","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,21]]}}}