Background: Normal voice production depends on the synchronized cooperation of multiple physiological systems, which makes the voice sensitive to changes. Any systematic, neurological, and aerodigestive distortion is prone to affect voice production through reduced cognitive, pulmonary, and muscular functionality. This sensitivity inspired using voice as a biomarker to examine disorders that affect the voice. Technological improvements and emerging machine learning (ML) technologies have enabled possibilities of extracting digital vocal features from the voice for automated diagnosis and monitoring systems. Objective: This study aims to summarize a comprehensive view of research on voice-affecting disorders that uses ML techniques for diagnosis and monitoring through voice samples where systematic conditions, nonlaryngeal aerodigestive disorders, and neurological disorders are specifically of interest. Methods: This systematic literature review (SLR) investigated the state of the art of voice-based diagnostic and monitoring systems with ML technologies, targeting voice-affecting disorders without direct relation to the voice box from the point of view of applied health technology. Through a comprehensive search string, studies published from 2012 to 2022 from the databases Scopus, PubMed, and Web of Science were scanned and collected for assessment. To minimize bias, retrieval of the relevant references in other studies in the field was ensured, and 2 authors assessed the collected studies. Low-quality studies were removed through a quality assessment and relevant data were extracted through summary tables for analysis. The articles were checked for similarities between author groups to prevent cumulative redundancy bias during the screening process, where only 1 article was included from the same author group. Results: In the analysis of the 145 included studies, support vector machines were the most utilized ML technique (51/145, 35.2%), with the most studied disease being Parkinson disease (PD; reported in 87/145, 60%, studies). After 2017, 16 additional voice-affecting disorders were examined, in contrast to the 3 investigated previously. Furthermore, an upsurge in the use of artificial neural network–based architectures was observed after 2017. Almost half of the included studies were published in last 2 years (2021 and 2022). A broad interest from many countries was observed. Notably, nearly one-half (n=75) of the studies relied on 10 distinct data sets, and 11/145 (7.6%) used demographic data as an input for ML models. Conclusions: This SLR revealed considerable interest across multiple countries in using ML techniques for diagnosing and monitoring voice-affecting disorders, with PD being the most studied disorder. However, the review identified several gaps, including limited and unbalanced data set usage in studies, and a focus on diagnostic test rather than disorder-specific monitoring. Despite the limitations of being constrained by only peer-reviewed publications written in English, the SLR provides valuable insights into the current state of research on ML-based voice-affecting disorder diagnosis and monitoring and highlighting areas to address in future research.

Dementia is a cognitive disorder that mainly targets older adults. At present, dementia has no cure or prevention available. Scientists found that dementia symptoms might emerge as early as ten years before the onset of real disease. As a result, machine learning (ML) scientists developed various techniques for the early prediction of dementia using dementia symptoms. However, these methods have fundamental limitations, such as low accuracy and bias in machine learning (ML) models. To resolve the issue of bias in the proposed ML model, we deployed the adaptive synthetic sampling (ADASYN) technique, and to improve accuracy, we have proposed novel feature extraction techniques, namely, feature extraction battery (FEB) and optimized support vector machine (SVM) using radical basis function (rbf) for the classification of the disease. The hyperparameters of SVM are calibrated by employing the grid search approach. It is evident from the experimental results that the newly pr oposed model (FEB-SVM) improves the dementia prediction accuracy of the conventional SVM by 6%. The proposed model (FEB-SVM) obtained 98.28% accuracy on training data and a testing accuracy of 93.92%. Along with accuracy, the proposed model obtained a precision of 91.80%, recall of 86.59, F1-score of 89.12%, and Matthew’s correlation coefficient (MCC) of 0.4987. Moreover, the newly proposed model (FEB-SVM) outperforms the 12 state-of-the-art ML models that the researchers have recently presented for dementia prediction.

Chronic obstructive pulmonary disease (COPD) is a severe condition affecting millions worldwide, leading to numerous annual deaths. The absence of significant symptoms in its early stages promotes high underdiagnosis rates for the affected people. Besides pulmonary function failure, another harmful problem of COPD is the systemic effects, e.g., heart failure or voice distortion. However, the systemic effects of COPD might provide valuable information for early detection. In other words, symptoms caused by systemic effects could be helpful to detect the condition in its early stages. The proposed study aims to explore whether the voice features extracted from the vowel "a" utterance carry any information that can be predictive of COPD by employing Machine Learning (ML) on a newly collected voice dataset. Forty-eight participants were recruited from the pool of research clinic visitors at Blekinge Institute of Technology (BTH) in Sweden between January 2022 and May 2023. A dataset consisting of 1246 recordings from 48 participants was gathered. The collection of voice recordings containing the vowel "a" utterance commenced following an information and consent meeting with each participant using the VoiceDiagnostic application. The collected voice data was subjected to silence segment removal, feature extraction of baseline acoustic features, and Mel Frequency Cepstrum Coefficients (MFCC). Sociodemographic data was also collected from the participants. Three ML models were investigated for the binary classification of COPD and healthy controls: Random Forest (RF), Support Vector Machine (SVM), and CatBoost (CB). A nested k-fold cross-validation approach was employed. Additionally, the hyperparameters were optimized using grid-search on each ML model. For best performance assessment, accuracy, F1-score, precision, and recall metrics were computed. Afterward, we further examined the best classifier by utilizing the Area Under the Curve (AUC), Average Precision (AP), and SHapley Additive exPlanations (SHAP) feature-importance measures. The classifiers RF, SVM, and CB achieved a maximum accuracy of 77 %, 69 %, and 78 % on the test set and 93 %, 78 % and 97 % on the validation set, respectively. The CB classifier outperformed RF and SVM. After further investigation of the best-performing classifier, CB demonstrated the highest performance, producing an AUC of 82 % and AP of 76 %. In addition to age and gender, the mean values of baseline acoustic and MFCC features demonstrate high importance and deterministic characteristics for classification performance in both test and validation sets, though in varied order. This study concludes that the utterance of vowel "a" recordings contain information that can be captured by the CatBoost classifier with high accuracy for the classification of COPD. Additionally, baseline acoustic and MFCC features, in conjunction with age and gender information, can be employed for classification purposes and benefit healthcare for decision support in COPD diagnosis. NCT05897944.

In addition to impairing the lung function, chronic obstructive pulmonary disease (COPD) also affects phonatory characteristics. Recent research highlights the potential of voice as a digital biomarker to support clinical decision-making. While machine learning (ML) can detect disease patterns from acoustic features, clinical relevance requires understanding the relationship between the disorder and acoustic features. This study investigates both statistical and clinical significance using Baseline Acoustic (BLA) and Mel-Frequency Cepstral Coefficient (MFCC) features with focusing on individuals with COPD and healthy controls (HC). Acoustic features derived from Swedish utterances of the vowel [a:], recorded via mobile phones from 48 age-matched participants (24 COPD, 24 HC; equal gender distribution), were analyzed. To reduce bias from varying recording counts, features were aggregated by averaging 10 randomly selected recordings per participant over 100 iterations. Vowel articulation was visualized in the vowel quadrilateral space using F1 (tongue height) and F2 (tongue advancement). Group differences were assessed using the Shapiro-Wilk test, Mann-Whitney U test (α = 0.05), Benjamini-Hochberg (BH) and Bonferroni corrections, Permutational Multivariate Analysis of Variance (PERMANOVA) test, and Cliff's Delta (δ). Of 101 features, 29 remained significant after BH correction and one after Bonferroni. Multivariate testing (p = 0.019) showed group separation. Additionally, 34 features demonstrated large effect sizes, suggesting potential as digital biomarkers. Voice data recorded via mobile phones capture meaningful acoustic differences associated with COPD. These findings support the integration of voice-based assessments into eHealth platforms for noninvasive COPD screening and monitoring, which is pending further validation on larger populations. NCT06705647.

Vowel-based voice analysis is gaining attention as a potential non-invasive tool for COPD classification, offering insights into phonatory function. The growing need for voice data has necessitated the adoption of various techniques, including segmentation, to augment existing datasets for training comprehensive Machine Learning (ML) modelsThis study aims to investigate the possible effects of segmentation of the utterance of vowel "a" on the performance of ML classifiers CatBoost (CB), Random Forest (RF), and Support Vector Machine (SVM). This research involves training individual ML models using three distinct dataset constructions: full-sequence, segment-wise, and group-wise, derived from the utterance of the vowel "a" which consists of 1058 recordings belonging to 48 participants. This approach comprehensively analyzes how each data categorization impacts the model's performance and results. A nested cross-validation (nCV) approach was implemented with grid search for hyperparameter optimization. This rigorous methodology was employed to minimize overfitting risks and maximize model performance. Compared to the full-sequence dataset, the findings indicate that the second segment yielded higher results within the four-segment category. Specifically, the CB model achieved superior accuracy, attaining 97.8% and 84.6% on the validation and test sets, respectively. The same category for the CB model also demonstrated the best balance regarding true positive rate (TPR) and true negative rate (TNR), making it the most clinically effective choice. These findings suggest that time-sensitive properties in vowel production are important for COPD classification and that segmentation can aid in capturing these properties. Despite these promising results, the dataset size and demographic homogeneity limit generalizability, highlighting areas for future research. Trial registration The study is registered on clinicaltrials.gov with ID: NCT06160674.

Background: As the older population grows, so does the prevalence of cognitive impairment, emphasizing the importance of early diagnosis. The Mini-Mental State Examination (MMSE) is vital in identifying cognitive impairment. It is known that degraded oral health correlates with MMSE scores ≤26. Objective: This study aims to explore the potential of using machine learning (ML) technologies using oral health and demographic examination data to predict the probability of having MMSE scores of 30 or ≤26 in Swedish individuals older than 60 years. Methods: The study had a cross-sectional design. Baseline data from 2 longitudinal oral health and ongoing general health studies involving individuals older than 60 years were entered into ML models, including random forest, support vector machine, and CatBoost (CB) to classify MMSE scores as either 30 or ≤26, distinguishing between MMSE of 30 and MMSE ≤26 groups. Nested cross-validation (nCV) was used to mitigate overfitting. The best performance-giving model was further investigated for feature importance using Shapley additive explanation summary plots to easily visualize the contribution of each feature to the prediction output. The sample consisted of 693 individuals (350 females and 343 males). Results: All CB, random forest, and support vector machine models achieved high classification accuracies. However, CB exhibited superior performance with an average accuracy of 80.6% on the model using 3 × 3 nCV and surpassed the performance of other models. The Shapley additive explanation summary plot illustrates the impact of factors on the model’s predictions, such as age, Plaque Index, probing pocket depth, a feeling of dry mouth, level of education, and use of dental hygiene tools for approximal cleaning. Conclusions: The oral health parameters and demographic data used as inputs for ML classifiers contain sufficient information to differentiate between MMSE scores ≤26 and 30. This study suggests oral health parameters and ML techniques could offer a potential tool for screening MMSE scores for individuals aged 60 years and older.