Assisted Intelligence in Medicine using
ES, version 2022-04-27 /
- Make medical knowledge better available for decisions
using interactive graph technologies. Particularly about interactive
differential diagnoses and recommendations of treatments,
i.e."meaningfull use" of available knowledge.
- A new vision proposed for the former GrApH-AI project.
- Roadmap ,
- A community of physicians, data scientists and software
- Given information from the patient and from medical knowledge, try
to provide recommendations.
- The complexity of both kinds of information require a
representation as graphs and the use of fuzzy logic.
- Make a synthesis of best current medical knowledge from different
sources as medical ontologies, courses, textbooks, litterature, and
above all medical experts from specialized scientific communities.
- Keep this synthesis up-to-date.
record as graph
- Patient information structured as a graph. Focus on the relations
between symptoms, problems and actions.
- In order to improve the knowledge base, analysis of large
populations of patient records. Improve the attributes of relations
and particularly their relative weights.
- Seek unsuspected patterns.
- Evaluate the results of treatments.
- Natural Language processing and conversion in graphs.
- Facilitate the understanding between graphs in human minds and
graphs in machines.
- Training of students playing with graphs in order to discuss
differential diagnosis and the potential benefits of next actions.
- Use cases
- A patient arrive with a problem, for example chest pain, what are
the likelihoods of possible problems and what are the relative
priorities of what should be done next?
which first questions ? which physical examination ? ask an ECG ?
ask lab tests ? order images ? begin a treatment ? decide if an
admission is necessary ? After every answer, re-evaluate the new
situation and adapt the visual graph.