HES-SO Genève

The European Union has recently introduced a suite of foundational digital regulations—the Cyber Resilience Act, the Artificial Intelligence Act, the Radio Equipment Directive, the NIS 2 Directive, and the Cybersecurity Act—that directly affect the engineering of software-intensive systems. While these instruments aim to enhance trust and security, their overlapping scopes generate a complex compliance landscape that software development must address at the design, implementation, and deployment stages. This paper examines the cross-regulatory impact of such EU cybersecurity legislation from a software engineering perspective, aiming to provide a set of guidelines and recommendations for implementing a compliance-by-design approach. We analyse and compare the five legal instruments, focusing on how their obligations intersect with each other. We then translate their regulatory requirements into actionable artefacts, ranging from architectural constraints and security controls to organisational processes, using a legal engineering approach. Finally, we propose a compliance-by-design lifecycle pattern that integrates regulatory alignment into requirements engineering, system design, and testing. To demonstrate applicability, we evaluate three representative use cases: an AI-enabled power plant, an autonomous drone delivery platform, and an AI-powered clinical decision support system. These examples demonstrate that multiple regulatory regimes often govern software-based systems. We conclude with practical recommendations for suppliers, deployers, and policymakers towards an integrated compliance framework to promote compliance-aware software engineering. Our findings indicate that the European digital landscape is shifting compliance from a post-hoc audit exercise to a design-time engineering principle. Embedding compliance early into the software development lifecycle not only supports regulatory alignment but also improves system resilience and trustworthiness.

We investigated the role of surface and structural similarities in the retrieval of everyday life events by means of a novel experimental paradigm in which participants could report any memories that came to mind while reading anecdotes instantiating familiar Schema-Governed Categories (SCGs, e.g., making up an excuse). Based on the idea that the availability of memories sharing both surface and structural similarities may conceal participants' ability to retrieve purely structural matches, we presented participants with short narratives describing associations of surfaces and structures which are either usually encountered (e.g., making up an excuse not to hang out with someone, Experiment 1) or less frequent in everyday life (e.g., making up an excuse not to go to an art exhibition, Experiment 2). The results from Experiment 1 replicated the traditional finding that when they are available, memories combining surface and structural similarities prevail over purely structural matches. However, Experiment 2 revealed that when a more direct competition was set between purely surface and purely structural matches, structurally based retrievals were the most frequent. Experiment 3 showed that target cues deprived of any surface features (e.g., making up an excuse) elicited more retrievals of usual than unusual surface/structure combos, suggesting the influence of association frequency in analogical retrieval. The data provide the first line of evidence suggesting that participants rely on highly familiar SGCs to process structural encodings and retrieve past experiences based on structural similarities. We discuss the implications of these findings regarding current theories of analogical retrieval holding that this process is predominantly guided by surface similarities.

Abstract Super-resolution localization microscopy (SMLM) has become a central tool for nanoscale biological research for its high spatial resolution and compatibility with wide-field microscopy. Achieving quantitative SMLM, however, requires homogeneous high-power illumination, nanometric axial stability, and precise multi-channel detection, features typically restricted to high-end commercial instruments or custom solutions in specialized laboratories. The cost of such microscopes and their technical complexity still limit the accessibility of these advanced imaging techniques. Several home-made single molecule microscopes and their submodules have been demonstrated as opensource, highly-customizable, and cost-effective alternatives for their commercial counterparts. Yet, implementation of such systems often requires expert knowledge in optics, electronics, and control system engineering. We introduce Open Blink, a compact open-source TIRF microscope integrating powerful homogeneous quad-line laser illumination, dual-channel detection, and active focus-lock stabilization for quantitative multi-color super-resolution imaging. Open Blink achieves a localization precision below 10 nm in dSTORM, supports a tunable, large field of view from 105 × 105 µm 2 up to 257 × 257 µm 2 , and maintains axial stability over hours, enabling high-throughput super-resolution acquisition. Built with predominantly off-the-shelf components, and full integration into the open-source software µ Manager where metadata registration ensures reproducibility, Open Blink offers a low threshold for adoption by easing implementation, use and maintenance. At a substantially reduced cost of approximately 70 000 Euros, among which the high-power laser combiner alone is less than 20 000 euros, Open Blink greatly improves accessibility for laboratories who wish to implement scalable high performance super-resolution microscopy based on single molecules.

Objectives: This study aimed to determine whether adults with tinnitus exhibit altered phase-specific mandibular kinematics during saliva swallowing and increased prevalence of tongue thrust and Eustachian-tube insufficiency versus tinnitus-free controls. Methods: This was a cross-sectional case–control study including adults with tinnitus and controls. Standardized computerized kinesiography recorded three spontaneous saliva swallows per participant. Primary outcomes were opening/closing time (OCT) and post-closure stabilization time (STT); total swallowing time (SWT) was secondary. Tongue thrust (TT) and tubal insufficiency (TI) were assessed clinically. Distributional assumptions were checked with Shapiro–Wilk; between-group comparisons used two-sided Mann–Whitney U tests and Fisher’s exact tests (TT, TI). Effect sizes included rank-biserial correlation (r), Hodges–Lehmann median difference (Δ), and odds ratios (ORs) with 95% confidence intervals. Co-occurrence of TT and TI and their relationships with OCT, STT, and SWT were evaluated within strata (cases vs. controls) using Fisher’s exact test, φ, Mann–Whitney U tests, and Spearman’s ρ. Given the marked imbalance in age and sex between groups, unadjusted non-parametric comparisons were complemented by multivariable models with adjustment for age and sex. An omnibus non-parametric combination test summarized case–control differences across OCT, STT, and SWT. Results: Statistical analysis was performed on 77 cases with tinnitus and 78 controls. Tinnitus cases showed longer OCT (1.75 ± 0.92 vs. 1.12 ± 0.62 s; p < 0.001; r ≈ 0.40; Δ ≈ +0.60 s) and STT (1.44 ± 0.88 vs. 0.84 ± 0.62 s; p < 0.001; r ≈ 0.42; Δ ≈ +0.60 s), while SWT differed modestly and was not significant (2.75 ± 0.69 vs. 2.57 ± 0.65 s; p = 0.115; r ≈ 0.15; Δ ≈ +0.18 s). TT was more frequent in cases (18.2%) than controls (6.4%; OR = 3.05, 95% CI 1.08–8.61; p = 0.029), whereas TI occurred in 16.9% of cases and 0% of controls (corrected OR = 32.85, 95% CI 1.92–563.49; p < 0.001). Within tinnitus cases, TT and TI did not show meaningful co-occurrence (φ ≈ −0.03; p = 1.00). TT+ tinnitus patients exhibited markedly prolonged OCT compared with TT− (median 2.22 vs. 1.45 s; Δ ≈ +0.88 s; r ≈ 0.60; p < 0.001), whereas STT and SWT were minimally affected; TI was not materially associated with any swallowing-time parameter. Spearman analyses confirmed a moderate monotonic association between TT and OCT in tinnitus cases (ρ ≈ 0.40; p < 0.001), with all other correlations small and clinically negligible. Age- and sex-adjusted analyses confirmed longer OCT and STT in tinnitus cases, whereas SWT remained non-significant; TT and TI also remained more frequent in cases after adjustment. The omnibus test indicated a clear global separation between groups across OCT, STT, and SWT (permutation p < 0.001). Conclusions: Adults with tinnitus exhibit a distinct swallowing signature characterized by prolonged OCT and STT, together with higher prevalence of TT and TI. TT in tinnitus patients is specifically linked to a pronounced prolongation of OCT, while STT and SWT remain largely unchanged, and TI shows no relevant impact on kinematic indices.

Person-centred care (PCC) has been increasingly promoted in wound management, yet its theoretical foundations and practical application remain unclear. This scoping review aimed to map and synthesise how PCC frameworks, concepts and outcome measures have been used in wound care. Following the Joanna Briggs Institute methodology and PRISMA-ScR guidelines, a systematic search was conducted across major databases for studies published between 2020 and 2025. Eligible sources included empirical research, reviews and conceptual papers addressing PCC in adults with chronic wounds. Data were extracted and analysed descriptively across conceptual and evaluative domains. Fourteen publications met inclusion criteria. Only one explicit framework of person-centred wound care was identified. Most studies referred to patient-centred rather than person-centred approaches and applied principles such as empowerment, shared decision-making and communication without consistent theoretical grounding. Outcome assessment focuses mainly on clinical or functional indicators, with limited attention to relational or experience-based dimensions of care. Some studies used the term person-centred as an unreflected keyword. Person-centred wound management remains conceptually fragmented, methodologically heterogeneous and sometimes unreflected. Greater theoretical precision, consensus on terminology and development of validated frameworks and measurement tools are required to translate person-centred principles into consistent, evidence-based clinical practice.

Dopamine transporter (DAT) SPECT is a validated biomarker for Parkinson's disease (PD) and related degenerative parkinsonisms. Interpretation relies on visual assessment supported by striatal image features such as striatal binding ratios (SBRs). Deep learning can aid this process but often underuses complementary data, lacks robustness to heterogeneous inputs, and offers limited interpretability. We built an end-to-end multimodal framework that encodes DAT images and scalar data (patient age and striatal image features) using, respectively, a vision transformer and a multilayer perceptron. A transformer-based fusion module then combined the encoded representations, while tackling possible missing inputs. Interpretability was provided through modality-level attention, spatial attention maps, occlusion analysis, and scalar feature saliency. Performance was evaluated on 664 Parkinson's Progression Markers Initiative (PPMI) cases and two local datasets A (N = 228) and B (N = 530) from different devices, including PD, atypical parkinsonisms, and non-degenerative subjects. Transfer learning involved pretraining on two datasets and finetuning on the third. On PPMI, the model reached 97.4% AUROC, 95.5% accuracy, 97.0% sensitivity, and 91.9% specificity, matching state-of-the-art performance. Results were similar on dataset B (98.6% AUROC) but lower on dataset A (92.6% AUROC), likely due to its smaller size and reduced image quality. Explainability analyses showed the model focused on clinically relevant striatal regions and identified key scalar features such as putamen SBR and asymmetry. The fusion module also supported stable predictions despite missing data. Our method efficiently combined multimodal data with heterogeneous datasets and partial multimodal data. Integrated explainability tools showed clinically meaningful attention that is expected to favor its adoption.

Ce document a t gnr automatiquement le

Healthcare systems contribute to about 5% of the emissions of the greenhouse gases globally when taken together. With 6.7% of its national emissions of greenhouse gases, the Swiss healthcare system is among the biggest contributors worldwide when taken per capita. Although isolated initiatives to reduce environmental impacts have emerged these last years, there is a strong need for better understanding the barriers and the possibilities for a heuristic framework like One Health to facilitate transformation, in particular in the Swiss context. The main objective is to explore to what extent One Health can facilitate transformation of the Swiss healthcare system and a secondary objective is to examine the role of bioethics in reducing healthcare's environmental impact. A focus group was conducted with nine experts from different disciplines (medicine, environment, veterinary medicine, public health) at the premises of the Swiss Medical Association in Bern on January 27th, 2025. The discussion lasted 180 min and was audio-recorded. Verbatim was transcribed and analyzed using thematic analysis following COREQ guidelines. Three additional individual interviews were conducted with participants who were not able to attend the focus group. Six key themes emerged: the need to reorganize healthcare delivery, the specificity of the Swiss governance system, the need to adopt solutions that are “good enough”, One Health as a heuristic framework for systemic change, the required expansion of the borders of bioethics to include environmental and animal welfare considerations and the possible tensions between different dimensions within the One Health framework. Integrating environmental perspectives through a One Health approach is a possible solution to reduce the environmental impacts of the Swiss healthcare system. The study identifies both barriers (complexity of the Swiss healthcare system, siloed departments) and opportunities (professional willingness to collaborate, the development of post-Covid initiatives) for implementing sustainable healthcare transformation in Switzerland.