SYMEXP
SYMEXP is a tool designed to enhance medical students critical thinking and help them with medical QA exams. Given a medical QA about diagnoses with correct (already known) and incorrect options, the goal of this system is to provide a set of explanations why the correct is correct, why other options are not and highlight missing or not detected informations. As a result the system provides template-based explanations that rely on the HPO ontology.
As soon as the question with set of options (both correct and incorrect) are provided, the explanations from the SYMEXP pipeline can be ‘generated’. Options fields are connected to HPO and the autocomplete is proposing existing diseases from the ontology, needed to provide relevant explanations.
Explanations can be enhanced using the finding converter option which, by default, will use the findings-to-symptom database to transform medical findings (temperature is 39°C) to symptoms (fever). If a OpenAI key is provided the system will enhance this database by predicting detected findings symptoms if they don’t exist in the database.
This work has been supported by the French government, through the 3IA Côte d'Azur Investments in the Future project managed by the National Research Agency (ANR) with the reference number ANR-19-P3IA-0002. This work has also been supported by the CHIST-ERA grant of the Call XAI 2019 of the ANR with the grant number Project-ANR-21-CHR4-0002.