Novel Deep Learning-Based Vocal Biomarkers for Stress Detection in Koreans
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Objective The rapid societal changes have underscored the importance of effective stress detection and management. Chronic mental stress significantly contributes to both physical and psychological illnesses. However, many individuals often remain unaware of their stress levels until they face physical health issues, highlighting the necessity for regular stress monitoring. This study aimed to investigate the effectiveness of vocal biomarkers in detecting stress levels among healthy Korean employees and to contribute to digital healthcare solutions.Methods We conducted a multi-center clinical study by collecting voice recordings from 115 healthy Korean employees under both relaxed and stress-induced conditions. Stress was induced using the socially evaluated cold pressor test. The Emphasized Channel Attention, Propagation and Aggregation in Time delay neural network (ECAPA-TDNN) deep learning architecture, renowned for its advanced capabilities in analyzing person-specific voice features, was employed to develop stress prediction scores.Results The proposed model achieved a 70% accuracy rate in detecting stress. This performance underscores the potential of vocal biomarkers as a convenient and effective tool for individuals to self-monitor and manage their stress levels within digital healthcare frameworks.Conclusion The findings emphasize the promise of voice-based mental stress assessments within the Korean population and the importance of continued research on vocal biomarkers across diverse linguistic demographics.
Abstract
Objective The rapid societal changes have underscored the importance of effective stress detection and management. Chronic mental stress significantly contributes to both physical and psychological illnesses. However, many individuals often remain unaware of their stress levels until they face physical health issues, highlighting the necessity for regular stress monitoring. This study aimed to investigate the effectiveness of vocal biomarkers in detecting stress levels among healthy Korean employees and to contribute to digital healthcare solutions.Methods We conducted a multi-center clinical study by collecting voice recordings from 115 healthy Korean employees under both relaxed and stress-induced conditions. Stress was induced using the socially evaluated cold pressor test. The Emphasized Channel Attention, Propagation and Aggregation in Time delay neural network (ECAPA-TDNN) deep learning architecture, renowned for its advanced capabilities in analyzing person-specific voice features, was employed to develop stress prediction scores.Results The proposed model achieved a 70% accuracy rate in detecting stress. This performance underscores the potential of vocal biomarkers as a convenient and effective tool for individuals to self-monitor and manage their stress levels within digital healthcare frameworks.Conclusion The findings emphasize the promise of voice-based mental stress assessments within the Korean population and the importance of continued research on vocal biomarkers across diverse linguistic demographics.
참고문헌 (43)
- Cohen S, "Psychological stress and dis ease", JAMA, 298, pp.1685~1687 (2007)
- Dimsdale JE, "Psychological stress and cardiovascular disease", J Am Coll Cardiol, 51, pp.1237~1246 (2008)
- Soung NK, "Psychological stress and cancer", J Anal Sci Technol, 6, pp.30~ (2015)
- Goodday SM, "Unlocking stress and forecasting its conse quences with digital technology", NPJ Digit Med, 2, pp.75~ (2019)
- Friganović A, "Stress and burnout syndrome and their associations with coping and job satisfaction in critical care nurs es : a literature review", Psychiatr Danub, 31(Suppl 1), pp.21~31 (2019)
- Kim JH, "The relationship between employee’s work-related stress and work ability based on qualitative literature analysis", J Ind Distrib Bus, 12, pp.15~25 (2021)
- Sharma S, "A comprehensive review and analysis of supervised learning and soft computing techniques for stress diagnosis in humans", Comput Biol Med, 134, pp.104450~ (2021)
- Cohen S, "A global measure of perceived stress", J Health Soc Behav, 24, pp.385~396 (1983)
- Roth AJ, "Rapid screening for psychologic distress in men with prostate carci noma : a pilot study", Cancer, 82, pp.1904~1908 (1998)
- Schultebraucks K, "Deep learning-based classification of posttraumatic stress disorder and depression following trauma utilizing visual and auditory markers of arousal and mood", Psychol Med, 52, pp.957~967 (2022)
- Paulmann S, "How psychological stress affects emotional prosody", PLoS One, 11, pp.e0165022~ (2016)
- Robin J, "Evaluation of speech-based digital biomarkers : review and recommen dations", Digit Biomark, 4, pp.99~108 (2020)
- Titze IR, "Principles of voice production" (2000), National Center for Voice and Speech
- Sondhi S, "Vocal indicators of emotional stress", Int J Comput Appl, 122, pp.38~43 (2015)
- Lu H, "StressSense: detecting stress in unconstrained acoustic envi ronments using smartphones", UbiComp’12: Proceedings of the 2012ACM Conference on Ubiquitous Computing, pp.351~360 (2012), Association for Computing Machinery
- Teager H, "Some observations on oral air flow during phonation", IEEE Trans Acoust Speech Signal Process, 28, pp.599~601 (1980)
- Ahamad A, "Evaluating AI system performance by recognition of voice during so cial conversation", 2022 5th International Conference on Contemporary Computing and Informatics(IC3I), pp.149~154 (2022)
- Krizhevsky A, "ImageNet classification using deep convolutional neural networks", 26th Annual Conference on Neural Informa tion Processing Systems 2012, pp.1~9 (2012), Institute of Electrical and Electronics Engineers (IEEE)
- Rumelhart DE, "Parallel distributed processing: explorations in the microstructure of cognition", pp.318~362 (1986), The MIT Press
- Desplanques B, "ECAPA-TDNN: empha sizes channel attention, propagation, and aggregation in a TDNN based speaker verification", INTERSPEECH 2020, pp.3830~3834 (2020)
- Hu J, "Squeeze-and-excitation net works", 2018 IEEE, 2018, 71327141~
- Dawalatabad N, "ECAPA-TDNN embedding for speaker diarisation", INTERSPEECH 2021, pp.35603564~ (2021)
- Gulati A, "Conformer : convolution-augmented transformer for speech recognition", INTER SPEECH 2020, pp.5036~5040 (2020)
- Zhu Y, "Self-atentive speaker embed ding for text-independent speaker verification", INTERSPEECH 2018, pp.3573~3577 (2018)
- Sheehan DV, "The Mini-International Neuropsychiatric Interview(M. I. N. I. ) : the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10", J Clin Psychiatry, 59(Suppl 20), pp.22~33 (1998)
- Yoo SW, "Valid ity of the Korean version of the Mini-International Neuropsychiatric Interview", Anxiety Mood, 2, pp.50~55 (2006)
- Schwabe L, "Chronic stress modu lates the use of spatial and stimulus-response learning strategies in mice and man", Neurobiol Learn Mem, 90, pp.495~503 (2008)
- Schwabe L, "Stress prompts habit behavior in humans", J Neurosci, 29, pp.7191~7198 (2009)
- Jacobsen PB, "Screening for psychological distress in ambulatory cancer patients : a multicenter evaluation of the Distress Thermometer", Cancer, 103, pp.1494~1502 (2005)
- Ma X, "The diagnostic role of a short scanning tool—the distress thermometer : a meta-analy sis", Support Care Cancer, 22, pp.1741~1755 (2014)
- Sousa H, "The clinical utility of the Distress Thermometers in non-oncological contexts : a scoping review", J Clin Nurs, 30, pp.2131~2150 (2021)
- Hellhammer DH, "Salivary cortisol as a biomark er in stress research", Psychoneuroendocrinology, 34, pp.163~171 (2009)
- Dressendorfer RA, "Synthesis of a cortisol-biotin conjugate and evaluation as a tracer in an immunoassay for salivary cortisol measurement", J Steroid Biochem Mol Biol, 43, pp.683~692 (1992)
- Flanagan JL, "Speech analysis, synthesis and perception" (1972), Springer Verlag
- Thienpondt J, "Integrating the frequency translational invariance in TDNNs and frequency positional informa tion in 2D ResNets to enhance speaker verification", INTERSPEECH 2021, pp.2302~2306 (2021)
- Xue J, "ECAPA-TDNN for multi speaker text-to-speech synthesis", 2022 13th International Symposium on Chinese Spoken Lan guage Processing (ISCSLP), pp.230234~ (2022)
- Kang W, "Deep-learning-based end-to-end spoken language identification system for domain-mismatched scenarios", 13th Conference on Language Resources and Evaluation (LREC 2022), pp.7339~7343 (2022)
- Kumawat P, "Applying TDNN architectures to analyze the duration dependencies of speech emotion recognition", INTER SPEECH 2021, pp.3410~3414 (2021)
- Yang B, "Embedding entities and relations for learning and inference in knowledge bases", The 3rd International Conference on Learning Representations (ICLR) 2015 (2015)
- Han H, "A deep learning-based stress-detection al gorithm with speech signals", Workshop on Audiovisual Scene Under standing for Immersive Multimedia 2018, pp.11~15 (2018)
- "Mental health and COVID-19: early evi dence of the pandemic’s impact"[URL]
- Murphy JK, "Needs, gaps, and opportunities for standard and e-mental healthcare among at-risk populations in the Asia-Pacific in the context of COVID-19 : a rapid scoping review", Int J Equity Health, 20, pp.161~ (2021)
- Vasudevan S, "Digital biomarkers : conver gence of digital health technologies and biomarkers", NPJ Digit Med, 5, pp.36~ (2022)