arXiv:2606.28798v1 Announce Type: new Abstract: Objective: ICD codes are central to reimbursement, research, and population health surveillance, yet automated coding systems often struggle to integrate diagnostic signals from both clinical narratives and structured electronic health record (EHR) variables. We evaluated whether frozen medical large language model (LLM) representations can serve as a shared embedding space for multimodal primary diagnosis category prediction. Materials and Methods: We constructed a MIMIC-IV cohort of 13,645 admissions from the 10 most frequent primary ICD-10 cod
The proliferation of medical large language models (LLMs) and accessible clinical datasets like MIMIC-IV enables advanced research into their practical healthcare applications.
This development could significantly enhance the accuracy and efficiency of medical coding, impacting reimbursement, research, and population health surveillance.
The ability to integrate multimodal diagnostic signals through LLM-based probing could lead to more robust and automated primary diagnosis category prediction systems.
- · Healthcare providers
- · Medical AI developers
- · Health insurance companies
- · Medical researchers
- · Traditional medical coders (some roles)
- · Inefficient healthcare billing systems
Improved accuracy and efficiency in medical billing and record-keeping through automated diagnosis prediction.
Reduced healthcare administrative costs and accelerated medical research by making vast datasets more systematically analyzable.
The development of a new standard for clinical documentation and diagnostic protocols, potentially shifting the skills required for medical professionals.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI