Force myography (FMG)-based wearable interfaces have emerged as promising solutions for non-invasive human–machine interaction (HMI) in assistive and rehabilitation applications. However, achieving reliable real-time gesture recognition using FMG is challenging due to the need for a complex sensor setup and computational algorithms. Moreover, the FMG system suffers from sensor shifting and improper force transmission. Therefore, this work proposes an affordable, multi-channel wearable FMG armband that reliably captures forces at multiple forearm muscle locations during different gestures. A real-time gesture recognition system classifies the FMG data from the armband into discrete instructions for assistive devices. An optimized four-channel sensor configuration for the armband was designed to capture distinct muscle deformation patterns associated with hand gestures. FMG signals were recorded from 10 subjects for their five different hand gestures. The recorded data underwent preprocessing, feature extraction and selection, and then machine-learning-based classification. Five classifiers, Random Forest (RF), Support Vector Machine (SVM), K-Nearest Neighbours (KNN), Extreme Gradient Boosting (XGBoost), and Multi-Layer Perceptron (MLP), were used to recognize these five gestures. The RF classifier achieved the highest offline classification accuracy of 92% and was subsequently implemented in real time on an ESP32 microcontroller unit (MCU). The system, based on an RF-trained model, depicted an accuracy of 81% in identifying real-time hand gestures. The implemented scheme was further tested to control the movement of an indigenous robotic arm using the predefined gestures. Such a simple, low-cost, and effective scheme based on FMG can help control prosthetic and robotic devices using human intentions.
Rabeta Yeasmin, Ezzuddin Abuhussein, A. R. Tubbs, Michelle Lee, Tomoko Fujiwara, Joel D. Bumgardner, Hongsik Cho, J. Amber Jennings
Long-chain ω-3 polyunsaturated fatty acids (ω-3 PUFAs), such as docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA), possess strong anti-inflammatory and antimicrobial properties, but their therapeutic potential is severely limited by high susceptibility to oxidative degradation. Covalent immobilization onto biopolymers using cleavable ester linkages offers a route to prevent oxidation while enabling stimuli-responsive payload delivery. In this work, a transesterification-based covalent conjugation approach was developed to graft DHA and EPA onto electrospun chitosan membranes (ESCM) to achieve controlled bioactive release tailored for wound healing. DHA and EPA were conjugated to ESCM via transesterification, and successful ester bond formation was verified using Fourier-transform infrared spectroscopy (FTIR). The degree of substitution and surface loading densities were quantified via High-Performance Liquid Chromatography (HPLC), X-ray Photoelectron Spectroscopy (XPS), and acid-base titration, while surface wettability was characterized by static water contact angle measurements. In vitro release kinetics were monitored over five days in neutral phosphate-buffered saline (PBS), acidic acetate-buffered saline (ABS), and PBS containing Pseudomonas lipase. Indirect cytocompatibility and cell migration were assessed using L929 mouse fibroblasts in transwell co-cultures evaluated with luminescence viability assays and live/dead fluorescence staining. Antimicrobial efficacy was determined against both planktonic and biofilm forms of Methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa. FTIR spectra confirmed successful ester linkage of DHA and EPA onto the electrospun chitosan nanofiber surfaces. The resulting functionalized membranes displayed a hydrophobic water contact angle of 127±1°, and quantitative HPLC analysis confirmed high surface substitution with approximate loadings of ~25 µg/cm 2 DHA and ~13 µg/cm 2 EPA. Elution assays demonstrated responsive release profiles, showing no measurable lipid leaching in neutral PBS, but pronounced release under acidic conditions (ABS) and enzymatically active conditions (PBS with lipase). In cell culture assays, the conjugated membranes showed high biocompatibility by maintaining greater than 90% L929 fibroblast viability while supporting cellular migration. Furthermore, the membranes significantly suppressed planktonic growth and biofilm formation of MRSA, although no significant inhibitory activity was observed against P. aeruginosa. Covalently conjugating DHA and EPA to ESCM resolves the primary clinical limitation of PUFA oxidation while preserving their therapeutic efficacy. By leveraging localized acidic pH and bacterial lipase activity as trigger mechanisms, the scaffold facilitates on-demand, targeted release of active lipids within hostile wound microenvironments. Supported by high cytocompatibility with dermal fibroblasts and targeted bactericidal action against MRSA biofilms, this composite membrane represents a viable bioactive dressing for the management of infected and chronic wounds.
Eve Sobirey, Jonte Schmiech, Matthias Bechstein, Fabian Flottmann, Maximilian Wagner, Jens Fiehler, Dieter Krause
The fabrication of realistic additively manufactured small vascular models for neurointerventional training remains challenging, as current manufacturing technologies often require a compromise between geometric accuracy and realistic catheter handling. Although several additive manufacturing technologies have been investigated, many fail to reliably reproduce patent lumina in vessels smaller than 2 mm. PolyJet printing enables the fabrication of highly accurate vascular geometries with open lumina; however, it produces comparatively rough inner surfaces, resulting in increased friction and unrealistic catheter handling, thereby limiting the tribological realism of the models. This study investigated whether surface coatings can improve the tribological performance of PolyJet vascular models while preserving their geometric accuracy. In this exploratory study, eight coating systems were evaluated using surface roughness measurements, cyclic catheter force measurements, short-term stability testing, micro-computed tomography, and clinical validation by experienced interventional neuroradiologists. Several coatings reduced surface roughness and catheter advancement forces compared with uncoated models. However, reduced roughness alone did not fully explain handling performance. Micro-computed tomography revealed substantial differences in coating homogeneity and lumen preservation, whereas force-path analyses demonstrated that homogeneous coatings produced smoother and more uniform force profiles. These objective findings were largely confirmed by the physicians’ subjective assessments. Among the investigated systems, Teflon spray provided the best overall balance between friction reduction, coating homogeneity, ease of application, and clinical acceptance, whereas Aquasure achieved the lowest catheter advancement forces but required a substantially more complex coating procedure. The results demonstrate that targeted surface modification can overcome one of the major limitations of PolyJet printed small vascular models by improving tribological realism while preserving their geometric accuracy.
Despite numerous efforts in electroencephalogram (EEG) signal processing and classification algorithms, predicting brain sensitivity during surgery remains a significant challenge. We refine here the checkpoint decomposition algorithm (CDA), a real-time approach for evaluating patient sensitivity during the early induction phase (first 10 min) of general anesthesia (GA), by segmenting the EEG signals into epochs of α -suppression (α S), where the α -band is absent, and iso-electric suppressions (IESs), characterized by low EEG amplitude. To predict brain sensitivity, the CDA incorporates three parameters: 1) the first occurrence time of iso-electric EEG traces, 2) the fraction of α -suppressions, and 3) their first occurrence time. We further develop a correlation approach based on the cosine similarity formula. Using logistic regression, we demonstrate that the three parameters computed in the CDA provide a reliable prediction of sensitivity, outperforming the Bispectral Index™ (BIS™), which was developed over 20 years ago for monitoring anesthesia depth. Sensitivity was defined by the total IES duration and the hypnotic (propofol) concentration when the BIS stabilized at 50. The BIS parameters extracted were the minimum value (nadir), the time to reach the minimum value, and the duration below 40. Finally, we found that the cosine correlation metric between the BIS values and the maximum α -band power frequency, α S proportion, IES, and α / δ power ratio reveals that lower BIS values do not consistently correlate with the presence of IES and α S. The checkpoint decomposition algorithm offers a valuable tool for evaluating and anticipating brain sensitivity to GA by providing complementary information compared to the BIS.
Paolo Oliva, Elisa Cereda, Elena Robles, Cecilia Cappelli, Simone Micalizzi, Alberto Bevilacqua, Andrea Rescalli, Nada Khaled Mansour
The rapid expansion of 3D printing in healthcare has enabled new paradigms in personalized medicine, particularly in the development of custom-made medical devices. However, in the European context, the integration of point-of-care additive manufacturing into routine clinical practice remains limited by the complexity of the Medical Devices Regulation (EU) 2017/745 (MDR), which poses significant challenges for hospital- and university-based laboratories in achieving regulatory compliance while preserving clinical integration. This study describes a real-world implementation of a hybrid organizational and regulatory model within the 3D Innovation Lab of Humanitas University (Milan, Italy), designed to bridge the gap between clinically embedded 3D printing activities and regulated medical device manufacturing. A top-down, regulatory-driven approach was adopted, based on the interpretation of MDR requirements and the implementation of ISO 13485 and ISO 14971 standards within a fully digital Quality Management System (QMS). The model integrates structured regulatory decision-making, formalized roles and responsibilities, and adapted design control, risk management, validation, and production processes tailored to a clinician-driven, custom-made workflow. The implementation resulted in ISO 13485 certification and formal registration as a manufacturer of custom-made medical devices under Annex XIII MDR. Key outcomes include: (i) the implementation of a regulatory decision framework enabling the systematic classification of in-house, custom-made, and CE-marked regulatory pathways within the broader point-of-care manufacturing landscape; (ii) the integration of clinical and manufacturing processes within a unified QMS, ensuring traceability and regulatory compliance across the device lifecycle; (iii) the ability to design, manufacture, and supply Class I and IIa custom-made devices to both internal and external healthcare institutions. Moreover, preliminary observations of this system have demonstrated clinical relevance in complex surgical scenarios, improving preoperative planning accuracy and enabling advanced patient-specific applications with a high precision surgical act, however comparative effectiveness data demonstration is still ongoing. These findings demonstrate the feasibility of combining clinical integration and regulated manufacturing within a single governance structure. The proposed hybrid model represents a structured and potentially replicable framework for advancing regulatory maturity in university-based 3D printing, supporting the transition from experimental point-of-care activities to sustainable, compliant medical device production.
