The proper use of standard classifications, such as the International Classification of Diseases (ICD) and coding of morbidity data has always been fundamental for all general epidemiological and many health-management purposes (WHO, 2016). One example is the use of the information flow of the Hospital Discharge Records (SDO) collected in national databases for monitoring hospitalization episodes provided in public and private hospitals and thus the provision of hospital assistance. This has become an indispensable tool for both administrative analyses (i.e., for accurate billing) and clinical elaborations (e.g., health quality assessment), which can bring to the planning of new measures to support healthcare and welfare activities or to more strictly clinical-epidemiological and outcome analyses. In this frame, although approaches to coding vary across institutions, clinical coding specialists frequently perform coding retrospectively. The assignment of codes to each patient episode of care during hospitalization is determined by different factors, among others by the coder's interpretation of the available case notes or the completeness of the electronic health records. As a result, accurate coding is dependent on both the intelligibility of the case notes and the coders' knowledge of medical terminology (Sundararajan et al. 2015). Several studies have indicated poor reproducibility of clinical coding (Tatham A., 2008) and poor accuracy which seems not dependent on the version of the standard coding system used, which in the case of SDO is ICD (Quan et al. 2014). In recent years, even if the application of artificial intelligence (AI) has begun to attract and, in some cases, assist clinicians in the practice of medical coding, the performances achieved by AI models do not meet expectations.

Effective human-robot collaboration hinges on robust communication channels, with visual signaling playing a pivotal role due to its intuitive appeal. Yet, the creation of visually intuitive cues often demands extensive resources and specialized knowledge. The emergence of Large Language Models (LLMs) offers promising avenues for enhancing human-robot interactions and revolutionizing the way we generate context-aware visual cues. To this end, we introduce SiSCo--a novel framework that combines the computational power of LLMs with mixed-reality technologies to streamline the creation of visual cues for human-robot collaboration. Our results show that SiSCo improves the efficiency of communication in human-robot teaming tasks, reducing task completion time by approximately 73% and increasing task success rates by 18% compared to baseline natural language signals. Additionally, SiSCo reduces cognitive load for participants by 46%, as measured by the NASA-TLX subscale, and receives above-average user ratings for on-the-fly signals generated for unseen objects. To encourage further development and broader community engagement, we provide full access to SiSCo's implementation and related materials on our GitHub repository.