A SMALL TONGUE MODEL-BASED END-TO-END COMPUTATION METHOD FOR REMOTE MEDICAL PRE-ASSESSMENT

Abstract

The invention relates to a data processing method for computer-assisted medical pre-assessment performed within an institutional local network. The method involves creating a structured case unit using clinical data, defining an uncertainty metric on this case unit, and comparing this uncertainty metric to at least one predetermined threshold value. Depending on the comparison result, the processing flow continues in the pre-assessment layer or is redirected to the clinical collaboration layer. In the clinical collaboration layer, the structured case unit is linked to other structured case units within a structured semantic memory using ontology-based common clinical concepts and semantic representation information. This linkage results in an explainable evaluation output.After the process is complete, the structured case unit is saved to the structured semantic memory, and the structured semantic memory is updated according to the defined update criteria. The invention relates to a computer-implemented medical pre-assessment method via a two-tiered data processing architecture that uses structured case units and structured semantic memory together.