This podcast discusses the current status and future projections regarding the use of large-scale language models (LLMs) to support the creation of referral letters and discharge summaries in the medical field. Examples from Japan and overseas, including OpTim, Infocom, and a demonstration experiment between Mie University and NTT, are introduced, and the importance of linking with electronic medical records and the approaches taken by new entrants (OCR technology, stand-alone applications, building local LLMs, etc.) are discussed. In particular, it is expected that electronic medical record manufacturers will lead the development of this field, and the possibility of commoditization in the near future is suggested.

Translated with DeepL.com (free version)

Ver todos los episodios