Illicit financial activities such as money laundering often manifest through recurrent topological patterns in transaction networks. Detecting these patterns automatically remains challenging due to the scarcity of labeled real-world data and strict privacy constraints. To address this, we investigate whether Graph Autoencoders (GAEs) can effectively learn and distinguish topological patterns that mimic money laundering operations when trained on synthetic data. The analysis consists of two phases: (i) data generation, where synthetic samples are created for seven well-known illicit activity patterns using parametrized generators that preserve structural consistency while introducing realistic variability; and (ii) model training and validation, where separate GAEs are trained on each pattern without explicit labels, relying solely on reconstruction error as an indicator of learned structure. We compare three GAE implementations based on three distinct convolutional layers: Graph Convolutional (GAE-GCN), GraphSAGE (GAE-SAGE), and Graph Attention Network (GAE-GAT). Experimental results show that GAE-GCN achieves the most consistent reconstruction performance across patterns, while GAE-SAGE and GAE-GAT exhibit competitive results only in few specific patterns. These findings suggest that graph-based representation learning on synthetic data provides a viable path toward developing AI-driven tools for detecting illicit behaviors, overcoming the limitations of financial datasets.
Recent advances in drug development and tissue engineering have emphasized the need for precise 3D-printed multilayered blood vessel biomodels. Although existing extrusion-based models integrate biomaterials and cells, they often lack geometric fidelity, limiting their practical use in tissue engineering. This study addresses a key constraint in the generation of multi-material cylindrical constructs by developing a new printing protocol to enhance trajectory generation for extrusion-based multi-material nested cylindrical model bioprinting process. To overcome this challenge, we have developed a computer-aided design and computer-aided manufacturing software capable of generating multi-material nested cylindrical models and their printing trajectory using an innovative printing protocol. The protocol strategically modifies the starting point of each layer and avoids collisions during fabrication. Furthermore, it optimizes the printing order to minimize tool changes and enables flexible adjustment of photopolymerization timing and pathing. We have tested this approach on three multi-material tissue models and compared the results with the models generated using the BIOCAD software (RegenHU). Findings demonstrate that our protocol significantly improves cylindrical structure integrity, minimizes printing collisions and reduces overall printing time. Comparative analysis confirms the superior printing fidelity of the tissue models printed using our method, validating its effectiveness for extrusion-based cylindrical bioprinting applications. This optimized trajectory-generation approach provides a robust framework for creating physiologically accuratein vitrovascular models, potentially accelerating drug discovery and reducing the reliance on animal experimentation in biomedical research.
This study addresses the challenge of creating datasets for cybercrime analysis while complying with the requirements of regulations such as the General Data Protection Regulation (GDPR) and Organic Law 10/1995 of the Penal Code. To this end, a system is proposed for collecting information from the Telegram platform, including text, audio, and images; the implementation of speech-to-text transcription models incorporating signal enhancement techniques; and the evaluation of different Named Entity Recognition (NER) solutions, including Microsoft Presidio and AI models designed using a transformer-based architecture. Experimental results indicate that Parakeet achieves the best performance in audio transcription, while the proposed NER solutions achieve the highest f1-score values in detecting sensitive information. In addition, anonymization metrics are presented that allow evaluation of the preservation of structural coherence in the data, while simultaneously guaranteeing the protection of personal information and supporting cybersecurity research within the current legal framework.
This study investigates the impact of marine salt accumulation on the Wells-type turbines of the Mutriku Wave Power Plant (MWPP). During the first half of 2020, a failure of the automatic freshwater washing system, combined with COVID-19-related mobility restrictions, resulted in an extended period without turbine cleaning. The information recorded by the data acquisition system (SCADA) in the MWPP (power provided by the turbines as well as vibrations) and ocean-state data were analysed to quantify the associated operational degradation.The results show that salt accumulation reduced the power relative difference of the affected turbine by an average of 46% (ranging from 31% to 59%) during the three-month failure period, and significantly altered the vibration response. For the sea-state conditions observed in 19th March 2020 (swh=2.0 m and mwp=13.4 s), the healthy-state vibration baseline ranged from 1.68 to 1.86 mm/s, whereas the measured average vibration on that day reached 4.25 mm/s. After manual cleaning, turbine performance rapidly returned to normal operating levels.These findings demonstrate that salt accumulation was the dominant factor affecting both turbine power output and vibration behaviour, highlighting the importance of effective cleaning strategies in Oscillating Water Column (OWC) energy converters.
Targeted Alpha Therapy (TAT) holds considerable promise for precision oncology, yet its clinical deployment in Europe remains limited. Beyond scientific considerations, progress is constrained by fragmentation across the radiopharmaceutical value chain, encompassing radionuclide production, radiochemistry, preclinical validation, clinical development, regulatory frameworks, infrastructure, and workforce training. This editorial focuses on astatine‑211 (211At) as a strategically important radionuclide for European targeted alpha therapy. Its favorable decay characteristics, chemical versatility, and theranostic potential make 211At a strong candidate for clinical translation. However, despite advances in radiochemistry and encouraging early‑phase clinical experience, current evidence remains largely preclinical or restricted to first‑in‑human studies, and broader clinical adoption remains uncertain. We argue that the main barriers are not isolated technical obstacles but a self‑reinforcing dynamic linking limited supply, restricted clinical evidence generation, regulatory complexity, and insufficient workforce readiness. Addressing these challenges sequentially or in isolation is unlikely to succeed. Instead, coordinated, value‑chain‑driven approaches are required, in which production, preclinical research, clinical evaluation, regulation, and training progress together. From a European perspective, initiatives such as ACCELERATE.EU: an EU-HORIZON-JU-IHI funded project bringing together a consortium of 17 partners across Europe with expertise in 211At and targeted alpha therapies, illustrate how distributed production networks, harmonized practices, co‑clinical development, and structured workforce training can contribute to greater autonomy, resilience, and equitable patient access. Establishing a decentralized and sustainable European supply model for 211At should therefore be considered a strategic priority to enable multicenter clinical trials, supply independence and support the future integration of targeted alpha therapy into routine clinical practice.
