2025
Autores
Taveira, É; Rêgo, S; Dutra, I;
Publicação
Abstract The digitalization of health care has accelerated the adoption of mobile health applications (mHealth apps) in Family and General Medicine in Portugal. These tools may support chronic disease management and clinical decision-making. However, limited high-quality scientific evidence and the absence of a national framework for certification and quality standards create uncertainty about their safe integration into clinical practice. This study aimed to characterize mHealth apps use among Family and General Medicine residents and physicians in Portugal. It also examined factors influencing app selection, barriers to adoption, and clinicians’ perceptions regarding the integration of Artificial Intelligence (AI) into clinical practice. An observational, cross-sectional, quantitative study was conducted using an online survey developed in LimeSurvey®. The questionnaire was distributed to residents and physicians registered in the Ordem dos Médicos® (Portuguese Medical Association) with active clinical practice. The final sample included 141 participants (73.8% female; 26.2% male). Data analysis used descriptive statistics, reporting absolute and relative frequencies. Most clinicians were aware of mHealth apps (97.9%), and 85.1% reported using them in clinical practice. Among users (n=120), 74.5% regularly used 2 to 5 apps. A total of 69 unique apps were identified, with 13 accounting for 63.0% of mentions, including Tonic®, UpToDate®, Cardio4all®, and PEM Móvel®. Apps were mainly used during clinical consultations (92.5%). The most frequent factors influencing app choice were ease of use (95.0%) and evidence-based clinical effectiveness (65.8%). Reported barriers included lack of knowledge about available apps (84.2%) and the absence of national evaluation standards (47.5%). Among non-users (n=21), the main structural barrier was poor integration with clinical information systems (71.4%). Regarding AI, 56.0% reported awareness of AI-integrated apps, mainly Tonic® and ChatGPT®. The same proportion considered AI use beneficial, especially for clinical decision support (80.9%) and administrative automation (62.4%). Key concerns included ethics, data security, privacy (74.5%), and limited interoperability. mHealth app adoption in Portugal is high but fragmented and largely driven by personal initiative (81.7%) and informal recommendations, with limited institutional guidance. Tonic was the only app identified by respondents as reporting compliance with ISO 13485 (medical software quality), ISO/IEC 42001 (AI management systems), and UEMS-EACCME clinical accreditation. Most clinicians perceive national regulatory guidance as insufficient (51.7%). Future progress requires urgent development of national framework for the curation and recommendation of mHealth apps aligned with international frameworks such as DiGA (Germany) and DTAC (England), increased digital health training, and improved interoperability with clinical systems to ensure safe, effective, and equitable use in Primary Health Care.
2026
Autores
Carrera, I; Criollo, J; Dutra, I;
Publicação
SMART TECHNOLOGIES, SYSTEMS AND APPLICATIONS, SMARTTECH-IC 2024, PT I
Abstract
This paper presents a novel approach to the computational representation of cellular lines using transformer-based embeddings. By leveraging state-of-the-art natural language processing techniques, we generate context-aware embeddings from biomedical literature from the PubMed database, offering a more nuanced and biologically relevant representation of cellular lines compared to traditional methods like TF-IDF and SVDD. We applied these embeddings to cluster cellular lines, using the elbow method to identify a set of distinct clusters that reflect biologically meaningful relationships. To evaluate the quality of these clusters, we employed the Topic Coherence metric, achieving a coherence score of 0.395, indicative of moderate consistency across clusters. The results demonstrate the potential of transformer-based models to improve drug discovery by identifying shared characteristics between cellular lines, enabling more accurate drug response predictions and advancing personalized medicine. This method offers an interesting improvement in the precision of cellular line modeling, paving the way for more efficient drug repositioning and targeted therapies in cancer research.
The access to the final selection minute is only available to applicants.
Please check the confirmation e-mail of your application to obtain the access code.