The episodes of Freezing of Gait (FOG) are a recurring symptom in people suffering from advanced stages of Parkinson's disease (PD). These are severe occurrences because they may cause falls to the patients, generating further traumas and concussions. In order to solve this yet ineffectively treated issue, this article describes the research that developed a device capable of predicting freezing episodes. On this project a wearable device was developed, which was able to predict freezing episodes based on the calculation of a freezing index (FI) determined by the signals obtained from an inertial measurement unit (IMU). This device was tested in three Patients and signals corresponding to normal gait and simulated Parkinson gait were taken. The results showed that FI obtained from Parkinson gait were much higher than those from a normal gait, validating this parameter as a key aspect in FOG prediction.
24 August 2026 in Methods in Ecology and Evolution
Identifying individual animals from vocalizations is an emerging research area in computational bioacoustics. This non‐invasive approach reduces reliance on physical capture and tagging for wildlife monitoring. Recent advances in this area leverage deep learning and bioacoustic foundation models adapted from species‐level classifiers. However, these models typically rely on fixed‐window inputs and may not fully capture temporal structure across extended or complex songs, which can contain information relevant to individual discrimination. Here, we evaluate whether modelling sequences of pretrained bioacoustic embeddings improves acoustic individual identification. We developed a framework that integrates transfer‐learned spectrotemporal representations from BirdNET with a lightweight long short‐term memory network. Unlike static baselines that use either the first embedding or an average over all embeddings, our approach processes time‐ordered embedding sequences, allowing the classifier to use information distributed across multiple windows. We evaluated the system using 87,865 vocalizations from 352 individuals across 7 species. We used four publicly available vocalization datasets with individual‐level labels, covering durations from 0.76 s (short calls) to 27.6 s (prolonged songs). Across five random seeds with stratified partitions, the framework achieved mean test accuracies between 93.9% and 98.3% and macro‐F1 scores ranging from 93.2% to 98.3%, without data augmentation. The clearest gains over the strongest static baseline were observed for the great tit and the great spotted kiwi, reaching and percentage points, respectively. Our results indicate that the contribution of temporal sequence modelling depends on vocalization structure, rather than providing uniform evidence that chronological order drives performance. The benefit of recurrent aggregation was limited or absent for short calls but evident for long vocalizations with multiple informative windows. By combining bioacoustic foundation models with lightweight recurrent modelling, this approach provides a scalable, CPU‐efficient tool for autonomous wildlife monitoring, particularly for species with extended or structurally complex vocalizations.
1. Identifying individual animals from vocalizations is an emerging research area in computational bioacoustics. This non-invasive approach reduces reliance on physical capture and tagging for wildlife monitoring. Recent advances in this area leverage deep learning and bioacoustic foundation models adapted from species-level classifiers. However, these models typically rely on fixed-window inputs and may not fully capture temporal structure across extended or complex songs, which can contain information relevant to individual discrimination. 2. Here, we evaluate whether modeling sequences of pretrained bioacoustic embeddings improves acoustic individual identification. We developed a framework that integrates transfer-learned spectrotemporal representations from BirdNET with a lightweight long short-term memory network. Unlike static baselines that use either the first embedding or an average over all embeddings, our approach processes time-ordered embedding sequences, allowing the classifier to use information distributed across multiple windows. 3. We evaluated the system using 87,865 vocalizations from 352 individuals across seven species. We used four publicly available vocalization datasets with individual-level labels, covering durations from 0.76 s (short calls) to 27.6 s (prolonged songs). Across five random seeds with stratified partitions, the framework achieved mean test accuracies between 93.9% and 98.3% and macro-F1 scores ranging from 93.2% to 98.3%, without data augmentation. The clearest gains over the strongest static baseline were observed for the great tit and the great spotted kiwi, reaching +1.3 and +2.8 percentage points, respectively. 4. Our results indicate that the contribution of temporal sequence modeling depends on vocalization structure, rather than providing uniform evidence that chronological order drives performance. The benefit of recurrent aggregation was limited or absent for short calls but evident for long vocalizations with multiple informative windows. By combining bioacoustic foundation models with lightweight recurrent modeling, this approach provides a scalable, CPU-efficient tool for autonomous wildlife monitoring, particularly for species with extended or structurally complex vocalizations.
Ventilator-associated pneumonia is acquired by many patients connected to an artificial airway, causing their death of a great percentage to those who are affected. A close relationship has been found between ventilator-associated pneumonia and hospital beds tilting angle which the patients rest on. The correct position has been established between 30° and 45° to reduce the likelihood of ventilator-associated pneumonia; so a tilt sensor linked to the hospital beds is a good method for measuring. The present article describes an electronic device which was developed able to create a link between an IMU sensor connected to a microcontroller and a Wi-Fi module that allows to display, in a real time graphical interface, located in the control center, the current position of hospital’s beds for being monitored by the nursing staff providing a record of the patients’ status in intensive care and emergency units. Device’s performance was checked by hysteresis and fixed position test along a previously defined time where errors in measurement of inclination remain within a suitable range.