Neonatal pain Neonatal pain or discomfort
Conditions
Interventions
The study device is the NeuroBell Luna wearable EEG system, a battery-powered, non-invasive neonatal EEG monitor designed for bedside use. The system records cortical electrical activity via adhesive
Sponsors
Universitair Medisch Centrum Groningen
Eligibility
Age
No minimum to 1 Years
Inclusion criteria
Inclusion criteria: Prospective CohortInfants may be enrolled if all the following apply:1. Admission to the UMCG NICU.2. Gestational age 24+0/7–42+0/7 weeks at the time of monitoring.3. Considered high-risk for pain, discomfort or neurological vulnerability, such as:a. Severe prematurity (<32 weeks),b. Postoperative status after major surgery,c. Suspected/confirmed neonatal encephalopathy (e.g. HIE, seizures, meningitis, intracerebral haemorrhage),d. Other conditions where EEG monitoring is clinically appropriate.4. Portable EEG monitoring deemed safe and feasible by the attending neonatologist.5. COMFORTneo Score will be available during monitoring, with at least one score recorded during the EEG period.6. Written informed consent obtained from parent(s) or legal guardian(s). Retrospective CohortRecords may be included if all the following apply:1. Infant was previously admitted to the UMCG NICU.2. aEEG/EEG data are available in the clinical archive.3. Gestational age at monitoring 24+0/7–42+0/7 weeks, comparable to the prospective cohort.4. COMFORTneo scores with timestamp are retrievable from the medical record (minimum one score required).5. Basic clinical context available (diagnosis, major interventions, sedation/analgesia history).
Exclusion criteria
Exclusion criteria: Prospective CohortInfants will be excluded if:1. Major congenital anomalies incompatible with life.2. Severe clinical instability where monitoring is unsafe or interferes with acute care.3. Skin breakdown, lesions or scalp abnormalities preventing electrode placement.4. Parents/legal guardians decline or withdraw consent. Retrospective CohortA dataset will be excluded if:1. No COMFORTneo Score documentation is available for linkage with EEG/aEEG.2. EEG/aEEG data are missing, incomplete or unusable due to artefact or file corruption.3. Documentation regarding clinical context is insufficient to support analysis.
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| The primary outcome concerns the feasibility of wearable EEG monitoring in the NICU and the successful generation of a synchronised dataset of EEG signals, COMFORTneo scores and physiological parameters.Feasibility will be assessed based on:successful acquisition of =2 hours of interpretable EEG recordingsEEG signal quality, including impedance levels, artefact burden and data lossavailability and temporal synchronisation of COMFORTneo scores with EEG recordingspractical feasibility of integrating wearable EEG monitoring within the clinical NICU workflowPrimary analyses will be descriptive. | — |
Secondary
| Measure | Time frame |
|---|---|
| Usability within the clinical workflowThe usability of wearable EEG monitoring in routine clinical practice will be evaluated among healthcare professionals involved in neonatal care (e.g. nurses and physicians). This will be assessed using the Dutch version of the System Usability Scale (SUS) to evaluate perceived usability and integration of the system within the existing NICU workflow.Dataset characteristicsThe quality and usability of the combined dataset will be evaluated based on:duration and completeness of available EEG/aEEG recordingssignal quality and proportion of usable EEG dataavailability of COMFORTneo scores with timestampsavailability of relevant physiological parameters and clinical eventsExploratory EEG analysesExploratory analyses will be performed to identify EEG characteristics potentially associated with pain, discomfort or neurological stress. The COMFORTneo score will be used as the clinical reference.Exploratory AI analysesThe combined prospective and retrospective dataset will be used for the development and exploratory evaluation of artificial intelligence (AI)-based methods for EEG signal analysis. EEG recordings will be segmented into short time windows and linked to clinical labels such as COMFORTneo scores and physiological parameters.These labelled data may be used for model training and validation using patient-level dataset separation or cross-validation approaches. The algorithms will be developed using Python-based scientific software environments, including libraries such as PyTorch, MNE and NumPy.All AI analyses will be performed offline on the collected research data and are intended solely for exploratory research purposes. The results will not influence clinical decision-making during the study. | — |
Countries
Netherlands
Contacts
Public ContactEMW Kooi
Universitair Medisch Centrum Groningen
Outcome results
None listed