2025/10/16 by Elena Bignami, Michele Russo · 1 voice
Engineering · Medicine · #Anatomy and Medical Technology #Artificial Intelligence in Healthcare and Education
paper · pdf · doi:10.1111/anae.70041
openalex created_date 2025/10/16 · openalex publication_date 2025/10/16 · openalex updated_date 2026/06/14
Ke et al. highlight the potential of generative artificial intelligence in streamlining peri-operative care and underscore the significant advancements in safety, consistency and clinician support [1]. The evaluation of a system that showed accuracy exceeding 97% reaffirms the growing role of artificial intelligence in modern clinical workflows. Generative artificial intelligence reinforces evidence-based and consistent care whilst reducing the likelihood of deviations and human error. Although the authors discuss the operational and safety benefits of their chatbot in the peri-operative setting, we believe the role artificial intelligence has in addressing individualised and population-level risks warrants further exploration. Predictive analytics complements generative artificial intelligence models such as those described. The integration of real-world data, analysed through machine and deep learning, could further enhance risk stratification and provide a more comprehensive understanding of individual patient needs and population health trends [2]. By strengthening predictive analytical algorithms, clinicians would gain the ability to: assess individual patient risks dynamically [3]; prevent clinical variations proactively; and support precision and personalised medicine goals, integrating population-wide data to inform tailored interventions for individuals. The relationship between generative artificial intelligence and predictive analytics underscores the possibility for individual patient data and dynamic risk prediction to enhance care quality. An example is the ‘tri-hybrid model’ which integrates generative artificial intelligence (guideline adherence and decision making); predictive analytics (identifying risk); and human intelligence (controllers and interpreters) [4]. This model would promote personalised care and ensure that clinicians remain in active control, providing an optimal balance of technology and human expertise. The potential to contextualise artificial intelligence systems to specific patient populations and healthcare systems cannot be overstated. By leveraging real-time feedback loops and iterative updates, such systems could be adapted to handle diverse clinical scenarios, ensuring their evolving relevance. Moreover, this proactive application of artificial intelligence enhances safety, reduces variability and offers clinicians a robust framework for managing complex decisions.