Authors Affiliation

Syed Taimoor Hussain Shah

Politecnico di Torino, Department of Mechanical and Aerospace Engineering, PolitoBioMed Lab, Corso Duca degli Abruzzi 24, Turin I-10129, Italy

Syed Adil Hussain Shah

GPI SpA, Department of Research and Development (R&D), Via Ragazzi del '99, Trento 38123, Italy

Andrea Buccoliero

Human Science Department, Università degli Studi di Verona, Lungadige Porta Vittoria, 17, Verona 37129, Italy

​

Iqra Iqbal Khan

Department of Computer Science, Bahauddin Zakariya University, Multan 60800, Pakistan

​

Department of Computing and Emerging Technologies, Emerson University Multan, Multan 60000, Pakistan

Syed Baqir Hussain Shah

COMSATS University Islamabad (CUI), Wah Campus, Department of Computer Science, Grand Trunk Road, Wah 47040, Pakistan

Angelo Di Terlizzi

GPI SpA, Department of Research and Development (R&D), Via Ragazzi del '99, Trento 38123, Italy

Giacomo Di Benedetto

7HC SRL, Rome 00198, Italy

​​​

 Received     2 June 2025           

 Accepted     24 June 2025             

 Online published  30  June  2025​​​​​​

ORIGINAL RESEARCH

Analyzing neonatal vocal expression: Methological approaches

to identifying neurological and psychiatric signatures

Publication:   Journal of Multiscale Neuroscience                DOI:  https://doi.org/10.56280/1703023560

Abstract

Analyzing neonatal vocal expression provides invaluable insights into brain function and the emergence of consciousness, as early vocalization patterns reflect neurodevelopmental trajectories and sensory integration processes. Despite progress in neonatal healthcare, identifying reliable neurological and cognitive markers from infant vocal sounds remains challenging, as it requires linking complex, multi-level brain activity with perceptual acoustic features. This paper reviews methodological approaches used to analyze neonatal vocal expressions, with a focus on techniques that bridge data-driven models with clinical applications. We examine computational methods, including signal processing, feature extraction algorithms, and machine learning models designed to capture vocal biomarkers of neurological or psychiatric disorders. Approaches include spectro-temporal analysis to detect atypical acoustic patterns, deep learning models like convolutional neural networks (CNNs) for automated feature learning, and explainable AI techniques that connect model outputs to clinically interpretable vocal features. We also explore multimodal approaches that combine vocal data with physiological and behavioral signals to improve diagnostic accuracy. The review addresses challenges in neonatal vocal analysis, including data scarcity, demographic variability, and the need for generalization across different recording environments. To mitigate these issues, we highlight advances in domain adaptation, transfer learning, and data augmentation, which enable models to generalize across diverse clinical scenarios. We emphasize the need for clinical validation and interdisciplinary collaboration to ensure practical adoption of these models in healthcare. Future research should focus on refining predictive models with larger, more diverse datasets and enabling real-time analysis for continuous neonatal monitoring. By evaluating existing methodologies and proposing future directions, this study aims to advance neonatal vocal analysis and support early diagnosis and intervention in pediatric healthcare.

​

Keyword: Neonatal vocal expression, Neurological and Psychiatric signatures, Signal processing, Machine/Deep Learning, Explainable AI, Pediatric healthcare

​

Conflict of Interest

The authors declare no conflict of interest

                
​Copyright: © 2025 The Author(s). Published by Neural Press.

This is an open access article distributed under the terms and conditions of the CC BY 4.0 license.

​

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, Neural Press™  or the editors, and the reviewers. Any product that may be evaluated in this article, or claim that made by its manufacturer, is not guaranteed or endorsed by the publisher.

 ​

References:

Ali, S., Tanweer, S., Khalid, S., Rao, N., 2021. Mel frequency cepstral coefficient: a review. In: Proceedings of the 2nd International Conference on ICT for Digital, Smart, and Sustainable Development, ICIDSSD 2020, 27-28 February 2020, Jamia Hamdard, New Delhi, India.

​

Alsalemi, A., Amira, A., Malekmohamadi, H., Diao, K., 2023. Novel domestic building energy consumption dataset: 1D timeseries and 2D Gramian angular fields representation. Data Brief, 47, 108985.

​

Al-Worafi, Y.M., 2024. Patient care related issues in the developing countries: monitoring parameters. In: Handbook of Medical and Health Sciences in Developing Countries. Springer, Cham, pp. 1–23.

​

Andonotopo, W., Bachnas, M.A., Dewantiningrum, J., Pramono, M.B.A., Stanojevic, M., Kurjak, A., 2025. AI and early diagnostics: mapping fetal facial expressions through development, evolution, and 4D ultrasound. Journal of Perinatal Medicine, 53, 263–285.

​

Arya, S.S., Dias, S.B., Jelinek, H.F., Hadjileontiadis, L.J., Pappa, A.-M., 2023. The convergence of traditional and digital biomarkers through AI-assisted biosensing: a new era in translational diagnostics? Biosensors and Bioelectronics, 235, 115387.

​

Gómez-Vilda, P., Gómez-Rodellar, A., Palacios-Alonso, D., Rodellar-Biarge, V., Álvarez-Marquina, A., 2022. The role of data analytics in the assessment of pathological speech—a critical appraisal. Applied Sciences, 12, 11095.

Guarrasi, V., Aksu, F., Caruso, C.M., Di Feola, F., Rofena, A., Ruffini, F., Soda, P., 2025. A systematic review of intermediate fusion in multimodal deep learning for biomedical applications. Image and Vision Computing, 158, 105509.

Hammoud, M., Getahun, M.N., Baldycheva, A., Somov, A., 2024. Machine learning-based infant crying interpretation. Frontiers in Artificial Intelligence, 7.

​

​