Mitsuteru Nakamura

Acoustic voice analysis demonstrates potential as a non-invasive biomarker for depression, yet its generalizability across languages remains underexplored. This cross-sectional study aimed to identify a set of cross-culturally consistent acoustic features for depression screening using distinct Vietnamese and Japanese voice datasets. We analyzed anonymized recordings from 251 participants, comprising 123 Vietnamese individuals assessed via the self-report Beck Depression Inventory (BDI) and 128 Japanese individuals assessed via the clinician-rated Hamilton Depression Rating Scale (HAM-D). From 6373 features extracted with openSMILE, a multi-stage selection pipeline identified 12 cross-cultural features, primarily from the auditory spectrum (AudSpec), Mel-Frequency Cepstral Coefficients (MFCCs), and logarithmic Harmonics-to-Noise Ratio (logHNR) domains. The cross-cultural model achieved a combined Area Under the Curve (AUC) of 0.934, with performance disparities observed between the Japanese (AUC = 0.993) and Vietnamese (AUC = 0.913) cohorts. This disparity may be attributed to dataset heterogeneity, including mismatched diagnostic tools and differing sample compositions (clinical vs. mixed community). Furthermore, the limited number of high-risk cases (n = 33) warrants cautious interpretation regarding the reliability of reported AUC values for severe depression classification. These findings suggest the presence of a core acoustic signature related to physiological psychomotor changes that may transcend linguistic boundaries. This study advances the exploration of global vocal biomarkers but underscores the need for prospective, standardized multilingual trials to overcome the limitations of secondary data analysis.

Physical pain, particularly musculoskeletal pain, negatively impacts the activities of daily life and quality of life of elderly people. Because pain is a subjective sensation and there are no standard assessment procedures to detect pain, we attempted to quantitatively determine the actual state of chronic pain caused by musculoskeletal organs and related factors based on questionnaires. First, we studied techniques for diagnosing diseases by monitoring the involuntary characteristics of the voice. Then, we applied the technique based on voice characteristics and proposed a voice index to detect chronic musculoskeletal pain. The voice index was derived based on the assumption that physiological changes due to chronic musculoskeletal pain also affect the vocal cords. Subjects in this study were adults, 65 years of age or older, with chronic pain in the musculoskeletal system (lumbar and/or knees). A large-scale population-based cohort study was conducted in 2019. Voice characteristics were extracted from the recorded voices of the subjects, and the characteristics with similar properties were organized into several principal components using principal component analysis. The principal components were further combined using logistic regression analysis to propose a voice index that discriminates between normal subjects and subjects suffering from chronic musculoskeletal pain. A discrimination accuracy of approximately 80% was obtained using the dataset corresponding to the participants with knee pain only, and a discrimination accuracy of approximately 70% was obtained during cross-validation of the same dataset. The proposed voice index may serve as a novel tool for detecting chronic musculoskeletal pain. Clinical impact: The voice-based pain detection holds clinical significance owing to its noninvasive nature, ease of administration, and potential to efficiently assess large populations within a short time frame.

We developed an algorithm for the early screening of autism spectrum disorder using baby cry sound. We collected the sound using a smartphone application to record it in their daily life. We employed the Japanese version of the Modified Checklist for Autism in Toddlers (M-CHAT) for the evaluation of the autism spectrum. We built a prediction model for binary classification using least absolute shrinkage and selection operator (LASSO) logistic regression. We evaluated the validity of the features as explanatory variables used in the model by cross-validation, and the accuracy of the file-wise prediction in the cross-validation was 0.63. This result proved the feasibility of the very early-stage screening of autism spectrum using baby cry.

Voice-based depression detection methods have been studied worldwide as an objective and easy method to detect depression. Conventional studies estimate the presence or severity of depression. However, an estimation of symptoms is a necessary technique not only to treat depression, but also to relieve patients’ distress. Hence, we studied a method for clustering symptoms from HAM-D scores of depressed patients and by estimating patients in different symptom groups based on acoustic features of their speech. We could separate different symptom groups with an accuracy of 79%. The results suggest that voice from speech can estimate the symptoms associated with depression.

In general, it is common knowledge that people’s feelings are reflected in their voice and facial expressions. This research work focuses on developing techniques for diagnosing depression based on acoustic properties of the voice. In this study, we developed a composite index of vocal acoustic properties that can be used for depression detection. Voice recordings were collected from patients undergoing outpatient treatment for major depressive disorder at a hospital or clinic following a physician’s diagnosis. Numerous features were extracted from the collected audio data using openSMILE software. Furthermore, qualitatively similar features were combined using principal component analysis. The resulting components were incorporated as parameters in a logistic regression based classifier, which achieved a diagnostic accuracy of ~90% on the training set and ~80% on the test set. Lastly, the proposed metric could serve as a new measure for evaluation of major depressive disorder.

It is empirically known that mood changes affect facial expressions and voices. In this study, the authors have focused on the voice to develop a method for estimating depression in individuals from their voices. A short input voice is ideal for applying the proposed method to a wide range of applications. Therefore, we evaluated this method using multiple input utterances while assuming a unit utterance input. The experimental results revealed that depressive states could be estimated with sufficient accuracy using the smallest number of utterances when positive utterances were included in three to four input utterances.