Leader: Alessandra Del Felice (UNIPD); Other collaborator(s): INRCA, CNR-ISTI. Margherita Bertuccelli (UNIPD)
We will design an experimental set-up to co-register electromyographic (EMG), angular motion through Inertial Motion Units (IMUs) and electroencephalographic (EEG) signals during movement in elderly. Neurophysiological and kinematic features of loss of balance, postural adjustments and upper limb motion trajectories will be identified. The neurophysiological correlates of movements and the classification of EEG, EMG and kinematic signals will be identified. Best stabilisation strategies identified during this step will be enacted by a rigid exoskeleton already in place @IAS Lab UNIPD. Ad-hoc psychometric testing and objective neurophysiological markers will provide information on embodiment and end- users acceptability of robot during the assisted tasks. Volunteers for data collection will be enrolled among community-dwelling and care home elderly.
Brief description of the activities and of the intermediate results:
Towards the completion of the experimental set up, we have concluded the fine-tuning of an ad hoc, controlled balance board to induce perturbation. We are now starting dry-runs before enrolling elderly subjects. Due to safety issues, we have also moved the experimental set-up to a different lab, with an in-build safety harness. We are in parallel pursuing the identification of neurophysiological correlates of embodiment: we concluded the first set of data acquisition and analysis of self-activated vs externally activated somatosensory stimulation during tasks with and without an exoskeleton. Preliminary analysis are in line with our hypothesis and confirm the reduced amplitude of SomatoSensory Evoked Potentials (SSEP) recorded from self-induced stimulation, which does not hold true while wearing a lower limb exoskeleton. Data collection is now continuing to increase the sample size for publication.
Lastly, we started enrolling participants for a study planned in collaboration with Age-IT colleagues from UNIFI - Design Faculty, aimed at providing end-users perspectives on exoskeletons using ad hoc questionnaires and focus groups
The collaboration with UNIFI-Design faculty is foreseen to provide practical guidelines for industries design/production of assistive robotic devices. Similarly, the identification for the first time of quantifiable biomarkers of embodiment of robots reinforces this output, in the perspective of empowering the Italian industrial sector in the field.
The experimental protocol foresees the use of a harness to ensure participants' safety. Data collection will thus take place in a fully equipped motion lab analysis. The setting will provide us with the opportunity to collect full set of data on a subset of volunteers during controlled balance perturbations including electroencephalography (EEG), electromyography (EMG), stereophotogrammetric motion capture, Inertial Measurement Units (IMU). The whole dataset will be published (i.e. Scientific Data) and will be, to our knowledge, the first open-access database of balance perturbation with EEG-EMG-IMUs-motion capture co-registration.
Please see the next reporting period.
The reference period has seen significant achievements in the project. Firstly, ethics committee clearance has been obtained for all planned experiments, except for the design/usability experiment in collaboration with the University of Florence (UNIFI).
A meeting was held at the Neurorobotics lab of the University of Padua (UNIPD) in person with prof. Carbone and Ing. Panizzolo in September 2024 to discuss a collaboration with the perspective of developing a soft lower limb exoskeleton to be triggered by the neurophysiological signals identified during balance perturbations.
In collaboration with the University of Calabria (UNICAL) and the cascade project partners 2SMARTEST and Moveo srl, the biomechanical features of a lower limb soft robot to stabilize lower limb during motion/falls in the elderly were defined. 2SMARTEST, in collaboration with prof. Carbone also has a development plan for a soft upper limb exoskeleton.
One major accomplishment was the completion of data collection for the experiment, identifying neurophysiological markers of exoskeleton embodiment. This experiment involved EEG recording during self- or external stimulation at the wrist during lower limb motion with and without the exoskeleton. The features of SomatoSensory Evoked Potentials (SSEPs) extracted from EEG signals provide a marker of embodiment based on the self-attenuation theory, where SSEPs have lower amplitude if the stimulation is self-induced. The results support our hypothesis that self-stimulation while wearing a lower limb exoskeleton provides a higher SSEP amplitude than without, confirming the non-complete embodiment of the system.
To deepen this area of research, a protocol on usability and design of lower limb exoskeleton was defined in collaboration with the Department of Architecture and Design at the University of Florence, which is part of Spoke 9. The UNIFI partner identified the validated questionnaire.
Another achievement was the completion of the perturbation paradigm on a controlled balance board. This technical advancement will enable data collection of kinematic and muscle data, and, in a subsample of participants, electroencephalography data during loss of balance.
Furthermore, an algorithm for biomechanical model reconstruction from IMUs positioned on SEINIAM guidelines EMG recording points was developed to optimize the experimental set-up. This in-house code will be compared with the commercial model provided by the IMU system, as the IMU and EMG full set-up is burdensome in terms of the number of sensors mounted on the participant and time for set up.
A joint effort with the group of prof. Mundo led to the collaboration and involvement of UNIPD in a MSCA Doctoral Network call (deadline 28th November 2024) promoted by UNICAL.
Lastly, initial discussions were also held on an agreement with an exoskeleton proprietary company (Uango, IT) for IPRs and knowledge sharing of outputs from the previous experiment.
We provide an objective biomarker of exoskeleton embodiment to be used during prototyping phases of lower exoskeleton development, with an impact both on scientific progress and industrial application.
The recruitment of elderly participants for lower- and upper-limb experiments is ongoing, with 60 healthy elderly volunteers enrolled to date.
Concurrently, a pipeline for assessing upper limb performance in elderly stroke patients and healthy controls has been developed and optimized using in-house MATLAB codes to identify functional motor primitives in upper limb tasks through comparison with video-based segmentation. An additional experimental setup exploring motor control through the integrity of cortico-spinal tracts to the upper limb has been developed, and data have been collected from three stroke patients and two healthy subjects as part of dry runs. Additionally, a data analysis pipeline has been developed in MATLAB to automatically identify Motor Evoked Potentials (MEPs).
Furthermore, dry runs, testing and optimizing balance board inclinations to induce significant perturbations have been run and provide a benchmark for further optimization. Negotiations with the exoskeleton proprietary company (Uango, IT) for IPRs and knowledge sharing of outputs from the experiment of exoskeleton control based on neurophysiological signals are continuing, with a strong involvement and collaboration of prof. Menegatti (WP leader) to provide the technical and engineering expertise.
Lastly, to advance the development of a prototype of a soft exoskeleton for lower limbs, based on the expertise of a SME (Moveo), which produces passive soft hip exoskeletons, and the background of a SME (2Smartest) on Memory Alloys Fibres as smart and innovative actuators, biweekly meetings are regularly held to strictly monitor progress and to collectively resolve issues arising in the process of designing and actuating
The main scientific advancement has been the conceptualisation of upper limb motor primitives and their automated identification from IMU signals, with the final aim of incorporating the concept of motor primitives in the control of robots’ motor patterns. This approach is pursued, in a concerted effort, by the Robotics group of UNIPD for lower limb robots.
The inclusion in the protocols of people with post stroke motor disability paves the way for applications also in rehabilitation and assistance. Widening the scope of our application to stroke significantly impacts on societal and health care costs, since in Italy stroke-related disability accounted for 871.000 DALYs (Disability-Adjusted Life Years) in 2021 (Feigin et al., 2024).
To increase the number of elderly participants for recruitment, seniors of a local daycare center were enrolled (OPSA, Opera Pia Sant’Antonio, Rubano, Padova).
In parallel, a pipeline for assessing upper limb performance in people with stroke and healthy controls has been optimized using MATLAB. Specifically, the in-house code performance has been improved by a Biomedical Engineering Master student for his dissertation in March 2025, to identify functional motor primitives in upper limb tasks.
Data collection is ongoing for the upper limb movement and motor tract integrity protocol, with data collected from six participants with stroke so far. Concurrently, a Biomedical Engineering Master’s student is working on the development of a data analysis pipeline to identify induced Motor Evoked Potentials (iMEP) in MATLAB.
Data collection of neurophysiological and kinematic data during perturbations is ongoing. Eight elderly participants and nine adult controls, with two of the adult controls also undergoing EEG co-registration, have participated in the experiment.
Collaborative efforts are also underway, with bimonthly online meetings held with the companies Moveo srl and 2SMArtEST S.r.l., winners of the cascade call of WP3 of Spoke 9, for the design and prototypization of a lower limb and upper limb soft exoskeleton with embedded smart actuators.
Practicalities of the collaboration with UNIFI, Department of Architecture and Design, to test on a large cohort of elderly usability of lower limb exoskeleton were defined. The major hampering factor is the specific expertise needed to run the experiments, with a UNIFI researcher needed @UNIPD for each data capture session and to run focus groups. At this point, due to insufficient funding for a dedicated researcher and a bioengineer to follow this project, with an estimated 4 person/months each, the following steps were not pursued.
Lastly, following a networking event at UNIPD, monthly meetings are conducted with Prof. Chiara Verbano of UNIPD, the local PI of Spoke 5 of PNRR AGE-IT, to implement clinical lean pathways for Upper Limb self-guided home-based rehabilitation based on findings of the technology-based determination of upper limb motor trajectory recovery.
Publications/achievement:
• Registration of protocol "Scoping Review on ipsilateral Motor Evoked Potentials as a biomarker for upper limb recovery after stroke" in Open Science.
• An amendment to the Ethics Committee has been submitted to authorize the use of a lower limb exoskeleton for follow up experiments
• Submitted to IEEE Transactions on Neural Systems & Rehabilitation Engineering: “Quantitative assessment of human exoskeleton integration through a neurophysiological marker of embodiment” (No: TNSRE-2025-00404), 26/03/25
• Baba Alfonc, Bulcão Visco Diego, Bertuccelli Margherita, Del Felice Alessandra “Il recupero motorio dell’arto superiore nei pazienti con ictus cerebrale: il ruolo dell’auto-riabilitazione domiciliare guidata. Una scoping review “.SIRN 2025, Parma, 10 aprile – Conference Poster
• Mariasole Pasinato, Margherita Bertuccelli, Stefano Tortora, Edoardo Trombin, William Tasinazzo, Emanuele Menegatti, Alessandra Del Felice and Giovanni Sparacino “Identification of neurophysiological markers of the embodiment of rehabilitation robotic exoskeletons” GNB 2025, Parma, 12 aprile – Extended Abstract
Scientific advancements are ensured by the development of the first automated analysis pipeline for identification and analysis of motor evoked potentials. An impact on Italian robotics industry is provided by the joint efforts in the Cascade call to produce a soft exoskeleton prototype with Smart memory alloy actuators, positioning the consortium at the forefront of this research field.
Lastly, the application of smart and lean management concepts to out-patients pathways is the first in the field, and might have health economic effects in terms of optimisation of care and resources. As an example, the estimated added value for the rehabilitation activities for a standard 20-day rehabilitation course increases by 46% with the lean pathway, with a net cost decrease of 168 euros for every 10 rehabilitation sessions and an activity added value of up to 212% (see Verbano et al., in preparation).
During the reference period, significant progress was made on data collection of neurophysiological and kinematic markers in upper and lower limb in elderly populations, both healthy and with neurological disorders.
Neurophysiological and kinematic features of upper limb movement were captured. Researchers conducted assessments on people with stroke in the acute phase, as well as follow-up evaluations at one- and three-months post-stroke. A total of 14 patients were enrolled in the study. Motor feature extraction algorithms, developed through two previous master's theses, were applied to analyze upper limb movements in healthy elderly and young adults samples (>20 participants).
Data capture of lower limb and balance loss was continued. We collected electroencephalography (EEG), electromyography (EMG), and kinematic data from both young (n=10) and elderly (n=10) participants. Researchers established an agreement with a care home to facilitate the recruitment and data collection of additional elderly participants, resulting in a total of 43 elderly participants at the time of reporting. The data collection process involved multiple full-day acquisition sessions at the care home. Preliminary data analysis on kinematic and EMG for the experiments was initiated in collaboration with the Information Engineering Department.
To enhance the recruitment of elderly participants, a meeting was convened at local seniors' daycare centers, specifically at the Opera Immacolata Concezione (OIC) in Padova, in July 2025. This initiative aimed to broaden the participant base for studies focusing on the elderly population.
Concurrently, the research team continued its work on upper limb recovery in people with stroke, conducting both clinical and neurophysiological data collection. Follow-up assessments were performed to track patient progress over time.
In an effort to translate research findings into practical healthcare improvements, meetings were held with Spoke 5 (led by Prof. Verbano). These discussions focused on developing performance measurement indicators to support Lean and Safety Management in territorial ambulatory healthcare settings, specifically for upper limb training.
The study on markers of instability saw continued progress with ongoing clinical and physiological data collection, including EMG and IMU recordings. Collaboration with a local Nursing Home facilitated the completion of recruitment for the remaining elderly participants, culminating in the finalization of data collection for a total of 50 elderly subjects.
Scientific publications
Dissemination Events