Throughout the years, many functional AI systems have been developed in our lab. Some exemplary examples are shown below.
Welzijn.AI is a new digital solution for continuously monitoring the mental wellbeing in the vulnerable elderly. Currently in the form of a mobile app, its ultimate aim is to be embedded within a furry cat toy through a speech-only interface. [ video ]
Additionally, we have developed various alternative prototypes to assess the different technological LLM approaches. For example, Roderick Leito investigated RASA's CALM approach for administering the EQ-5D: video.
An extended English demo video of Wellbeing.AI has also been prepared by Bram!
The STRIP Assistant (STRIPA) is a stand-alone, web-based software tool that was used to perform a pharmaceutical analysis, an important step of the STRIP process. Data on diagnoses and current drug use (collected via SHiM and the actual medical record), recent measurements and laboratory values (e.g. renal function, blood pressure) and possible adverse drug reactions, as listed in the patient’s medical record and according to patient information (SHiM) were entered in STRIPA. The assignment of drugs to diseases has been implemented through a drag and drop mechanism. [ talk | paper1 | paper2 | video ]
SNPcurator is a system for information extraction specifically in the genome wide and candidate genes studies, constructed out of different natural language processing (NLP) modules to aid scientists in their search for relevant disease-associated SNPs through an intuitive web interface. It incorporates both syntactic and semantic methods to extract relevant information from PubMed abstracts such as cohort size and ethnicity, SNP ids and the reported results. [ paper ]