The invention discloses an intelligent auxiliary decision-making method and
system for multi-
modal data fusion of
multiple myeloma, and relates to the technical field of
medical information, and the method comprises the steps: obtaining multi-source clinical data and authoritative literature of a patient, and building a literature
database after preprocessing; configuring an interface to automatically capture new literatures, and generating a dynamic
knowledge base in combination with semantic conflict detection,
knowledge extraction and expert auditing; according to the
knowledge base, fusing the multi-
modal clinical data by adopting federal learning to generate a fusion
feature matrix; constructing a causal relationship network, combining matrix prediction and generating a causal evidence network, and grading the causal relationship; and associating the
knowledge base with the evidence network, generating an interpretable report containing a prediction result and the like, and realizing full-process
traceability. Through three designs such as automatic
knowledge acquisition and updating, the static
lag problem of the knowledge base is broken through, and the medical frontier is synchronized; a
causal reasoning system is constructed, defects of a traditional model are overcome, transparent reasoning is achieved, and the credibility of doctors is improved.