Incoming Research Mobility Grant
Incoming Research Mobility for Research Staff from Foreign Universities and Centres, University of Deusto, Spain.
- Sara Lenzi, Ikerbasque Research Fellow at DeustoTech – Instituto Tecnológico Deusto
- PerMagnus LIndborg, Visiting Researcher (City University of Hong Kong)
Identifying Harmful and Annoying Sound Sources in the Acoustic Environment at a Neonatal Intensive Care Unit
Having a prematurely born baby taken into a neonatal intensive care unit (NICU) is an extremely stressful experience for both the infant patient and their caretakers. The optimal functioning of life support, such as respiratory support systems and medical interventions, are of primary importance. Of secondary importance is, one could say, all the rest: comprehensive environmental safety, continuous emotional support, immense patience, and perhaps – luck. The present study focuses on one of the most important environmental stressors: noise. While it is well known that noisy acoustic environments at hospitals negatively impact patients, families, and healthcare staff (de Lima Andrade et al., 2021; Özcan et al., 2019), it might appear intractable. Noise, per definition, is an ‘unwanted side effect’ from some other action or event, which might be of primary importance, yet improvable. More than any other environmental stressor, noise tends to escape our attention yet has insidious long-term effects.
Field studies have evidenced noise levels that are considerably higher (Smith et al. 2018) than what. One study found that the recommended maximum level of 45 dBA was exceeded 90% of the time – day and night (Mayhew 2022). Noise is a major stressor for premature infants. The most common indicators of physiological stress and pain include changes in heart and respiratory rates, blood pressure, oxygen levels (tcP02) and saturation (O2 Sat), intracranial pressure, vagal tone, skin blood flow, and palmar sweat (Peng et. al, 2009; cited in Knutson 2012 p. 9). Excessive and unfamiliar sounds can disrupt vital signs, sleep, and behavior, and may cause long-term developmental issues, including delayed language acquisition (Boyar and Gennattasio, 2023. We are also concerned with how noise influences caregivers, in particular the mother and father. Their psychological wellbeing certainly mediates the long-term health of the baby. Ideally, the premature baby’s soundscape should be consistent with the intrauterine environment (cf. Knutson 2012 p. 11). We may understand the acoustic environment at a NICU in layers – incubator, caregiver, room, ward – each with its own set of sound sources: some beneficial, others detrimental. We may distinguish between: Continuous background (machines, HVAC, etc); On-off midground (various sources including distant voices, foot steps), and; Occasional foreground (alarms with differing degrees of urgency, close voices, crying, furniture).
Our previous study (Lenzi et al., 2024) lent evidence to the common observation that staff conversations and activities are significant contributors to nighttime sleep disturbances Although staff training can help (Carvalhais et al., 2015), many nurses are unaware of their contribution to noise and feel powerless to effect change (Spagnol et al., 2022). Traditional noise-monitoring technologies rely on sound pressure level (SPL), which is not easily understood by non-experts and fails to capture the emotional quality of specific sound types. Machine Listening (MLi) methods offer promising alternatives but have focused mainly on outdoor settings due to privacy concerns in hospitals (Lenzi et al., 2023). As a result, there is a critical gap in developing MLi tools to assess and improve indoor soundscapes, particularly in NICUs. A sustainable, human-centered, automated solution to monitor and predict the emotional and acoustic quality of these environments is urgently needed to support better healthcare outcomes and working conditions (Lindborg 2025 p. 31-38).
The University of Deusto is launching a new line of research on the impact of sound on premature babies in the context of Intensive Care Units (ICU), towards the design and implementation of Machine Learning and design-driven interventions. The research is kick-started in collaboration with the Neonatal ICU of the Hospital de Donostia-San Sebastián which will grant access to data collection. The research aims to create an integrated model of the acoustic environment in Neonatal Intensive Care Units (NICUs), focusing on how various sound sources—such as mechanical devices, human activity and voices, and medical alarms—affect the physiological and emotional well-being of premature infants. The core objective is to raise awareness among healthcare professionals, enabling them to understand and mitigate the negative impact of poor soundscapes on patients. Ultimately, this research will support the development of a novel Machine Listening (MLi) application capable of automatically monitoring and evaluating the acoustic quality of NICUs.
Recent relevant publications
- Lenzi S, Lindborg PM, Spagnol S, Kamphuis D, Özcan E. (2024/06). “Perceived quality of a nighttime hospital soundscape” Noise Mapping, vol. 11, no. 1, 2024, pp. 20240010. https://doi.org/10.1515/noise-2024-0010
- Lindborg PM, Lenzi S, Li SX, Li X (2025/01). “SleepSound: Charting the Nighttime Soundscape and Sleep Quality in Hong Kong through Machine Listening and AI-supported Information Design”. In Pauletto, S. (2025, January 29) Proceedings of the 3rd Conference on Sonification of Health and Environmental Data (SoniHED 2025, https://soundforenergy.net/sonihed2025). Stockholm, Sweden. https://doi.org/10.5281/zenodo.15032831, p. 30-37.
- Sara Lenzi, PerMagnus Lindborg, Ningze Han, Simone Spagnol, Daan Kamphuis, Elif Özcan (2023/09). “Disturbed Sleep: Estimating Night-time Sound Annoyance at a Hospital Ward”. Forum Acousticum, 10th Convention of the European Acoustics Association 2023, Torino, Italy.