1. Introduction

The Special Interest Group of the Japan Association for Medical Informatics the “Electronic Medical Record Research Group” [8], has been studying methods for exchanging medical data between different facilities since 1994.

In 1995, the idea of data exchange with attributes was born and was then put into practice with Standard Generalized Markup Language (SGML). This standard was named Medical Markup Language (MML). The group members took part in the Electronic Medical Chart Research Project of the ex-Ministry of Health and Welfare, launched in 1995. During the three year period of their involvement in this research and development, the exchange standard was improved to a practicable level. To handle data that cannot be expressed in MML(eg. images), a method of referring to external files such as DICOM from MML instances was established [2, 3] and, at the same time, work to detail MML to a practical level of implementation was carried out in cooperation with an ex-Ministry of Health and Welfare working group (with Professor Kimura at Hamamatsu University School of Medicine as the leader) to formulate the operating policy for medical information exchange such as how to combine DICOM, MML, HL7, etc.

Along with some practical movements in the implementation of MML, the need for specialized structures for various medical fields became evident, but it was in fact impossible for this group to cover all medical fields. Besides, considering the influence over the entire structure by new designs/changes of partial structures that are expected to occur frequently, it was not efficient, with respect to version control and the like. Therefore, a method to modularize MML with the XML Namespace suggested by W3C in March 1999 and combine modules as necessary was adopted. In this way, it became possible to propose descriptive formats particular to each medical field, and to divide logical structure development work procedures. This specification and relevant data are open/controlled at the MedXML site [1].