FR2781297A1 - Procedure for generation of code files in medical database systems, where data is stored in different classification systems - Google Patents

Abstract

The method involves generation of a code file from a database, in which the data is stored in several classification systems (CL1,CL2). The code file (FC) is generated from several items listed in a natural language by the database users. Each item has a classification code (C1,C2) attached to it, according to its classification system, so that the codes yield connections (LD1,LD2) to other databases. An Independent claim is included for a code file generation system.

Description

- 1 "Method and system for generating a coding file, and

  application to medical information systems "

DESCRIPTION

  The present invention relates to a method for generating a file. It also relates to a system implementing this method, as well as its application in particular to medical information systems. In the field of scientific and more particularly medical information, databases have become the essential tools and vectors of access to

  information. This information is necessarily subject to classifications and codifications allowing identification and automated access to the information sought.

  However, the users of these databases, who most of the time are not specialists in classification and coding systems, often encounter difficulties in accessing information due to the ignorance of the tree mechanisms attached to these systems. . This generally results in a loss of

time and efficiency.

  This problem is aggravated when it is a scientific or technical field in which several classification systems coexist. We can notably cite the case of medical databases which must currently take into account a transition between the WHO classification (CIM9) and the new WHO classification (CIM10). One can also cite the field of patents in which at least two classification systems coexist, namely the international system (WIPO) of patent classification, and the American classification system. The object of the invention is to remedy these drawbacks by proposing a method for generating a 2- codification file which allows a user of this file to evolve in a semantic universe and a natural language which are familiar to him without require him to know and use classifications and codifications. The aforementioned objectives are achieved with a method for generating a codification file in an information space, in particular medical information, stored in at least one database and which can be classified according to a plurality of classification systems. According to the invention, the coding file is generated by accumulation of items written in natural language, and to each item within the coding file, several classification codes are assigned, each belonging to a classification system among the plurality of systems. classification, so the classification codes so assigned provide links to other

  features or databases.

  Thus, a user of the codification file, by simple access to a diagnostic wording written in natural language, is given multiple possibilities of access to information or functionalities thanks to the links provided by the classification codes assigned to this wording. , without this user needing

know codes.

  In an advantageous embodiment of the method according to the invention, each item within the coding file is also assigned a unique code constituting the smallest common part between this item and the different classification codes assigned to said item. This unique code is representative of the unique character of each item within the coding file. These -3 items, made up of labels written in natural language, are not processed or filtered but preferably kept as they are. Preferably, the unique code attached to each item within the coding file is assigned sequentially.

  It should be noted that each label is itself unique in the file offered to the users' choice.

  The unique code is just one of 10 qualifiers of this unique label. This unique code allows a simple link with other characteristics including, for example, characteristics of previous observations which qualify the wording by notions of frequency and origin. It is important to note that the codification file thus obtained comes from a work of collection and processing upstream, then is offered at the choice of a user. The latter chooses from this coding file one or more items meeting his expectations which are stored in a file resulting from the

user choice.

  The method according to the invention can also advantageously be used to provide transcoding between a first classification system and a second classification system from among the plurality of classification systems via the codification file generated with

this process.

  In a practical form of implementation of the method according to the invention, it is arranged to generate, for each line of the coding file, a label written in natural language and constituting the item, a unique code attached to this label, and one or more codes

one or more classifications.

  According to another aspect of the invention, a system is proposed for generating a coding file in an information space, in particular medical information, stored in at least one database and which can be classified according to a plurality of classification systems, implementing the method according to the invention. This system is characterized in that it includes means for accumulating within this codification file items written in natural language by users of this database, and means for assigning to each item within said codification file a plurality of classification codes each belonging to one of the plurality of classification systems, so that the classification codes so assigned provide

  links with other functionalities or databases.

  In an advantageous version of the system according to the invention, it further comprises means for assigning to each item within the coding file a unique code constituting the smallest common part between this item and the different codes of

  classification assigned to said item.

  The system may further include means for transcoding between a first classification system and a second classification system from among the plurality of classification systems via the

  codification file generated with this process.

  The method and the system according to the invention find a preferred application in the field of medical information. Each item then consists of a description of diseases, diagnoses or reasons for medical consultation, and the codification file provides in particular access to databases and functionalities which use or contain at least one of the classifications used in said codification file. or in a file resulting from a choice

  carried out within said coding file.

  Other features and advantages of the invention will appear in the description below. To the

  attached drawings given by way of non-limiting examples: 5 - Figure 1 illustrates the essential steps of the method for generating a coding file according to the invention; - Figure 2 illustrates a mode of use of a codification file generated with the method according to the invention, with access by dynamic links to external functionalities; and - Figure 3 illustrates a transcoding operation between two classifications performed with

the method according to the invention.

  We will now describe, with reference to FIG. 1, the essential steps of the method for generating a coding file according to the invention, in the

medical information.

  A FC coding file according to the invention consists of an accumulation of diagnostic texts freely written in natural language by doctors using their own terms. These labels may have been entered after interviews, surveys, collection campaigns or directly by doctors in their daily practice. The descriptive labels constitute the items of the FC coding file. By way of nonlimiting example, the following labels may be encountered:

- SPINOCELLULAR LIP CARCINOMA

- MALIGNANT TUMOR LIPS

- MALIGNANT TUMOR BASE LANGUAGE

- MALIGNANT TUMOR GLAND PAROTIDE

  etc. It is important to note that the different items or labels can be accumulated sequentially, without

  6- filtering correction or prior storage.

  For each label LI having been the subject of an entry SL, a first coding operation COl is carried out according to a first classification system5 CLl, - for example the 9th revision of the WHO (World Health Organization) -, and a second classification operation C02 according to a second classification system CL2, - for example the 10th revision of the WHO. These two coding operations lead to the allocation of two classification codes C1, C2 for the wording concerned. In addition, an AC allocation operation

  a unique DFCC code with the relevant label is made.

  This unique code, designated by way of nonlimiting example under the term of. "Diagnosis Fixed Continuity Code

  (DFCC) ", is preferably assigned sequentially.

  But we can of course in the context of this

  provide for other forms of attribution of the single code.

  At the end of these coding and attribution operations, there is a series of items provided with classification code and a DFCC code which constitute

the FC coding file.

  This codification file can of course be updated by entering new items or new codifications depending on the evolution of

  classification systems in the area concerned.

  It should be noted that the FC codification file, which is initially offered at choice, has

  characteristics resulting from the modalities of its constitution.

  It can be accessed by the wording itself or by using the DFCC code. These characteristics are among others frequency or origin characteristics, and allow

for example:

  - to use or to constitute tables of frequencies and origins of these items, - to modulate the use of the file according to the data of frequency and origin, - to modulate the use of tables in transcoding by these notions of frequencies in case of type relation

  "l: n" between classification codes CL1 and CL2.

  By way of nonlimiting example, an extract from a codification file is given below: DFCC Classification 1 Classification 2 Description Diagnosis

  596 140-9 C00.9 SPINOCELLULAR LIP CARCINOMA

  597 140-9 C00.9 MALIGNANT TUMOR LIPS

  598 141-0 C01.0 MALIGNANT TUMOR BASE LANGUAGE

  599 141-9 C02.9 MALIGNANT TUMOR LANGUAGE

  We will now describe a mode of use of a FC coding file by a doctor in his practice, with reference to FIG. 2. We consider that this doctor has a microcomputer in which a set of software is installed. and utilities, such as office management software, communication software with social security and mutual insurance organizations, diagnostic assistance software, or even

  software to help prescribe drugs.

  Some of these software or utilities refer to diagnostic classification codes and sometimes require the entry of these codes. Thanks to the coding file obtained with the method according to the invention, a doctor who uses these software is not required to know the different classification systems, since the user of the coding file makes it possible to generate dynamic links with functionalities.

  requiring classification codes.

  Indeed, the doctor accesses the codification file and enters a request expressed in language

  natural, for example, "Malignant tumor on a lip".

  An automated search procedure allows him 8 - to directly access the wording closest to his request. But it can also be foreseen that this research is carried out using keywords according to conventional techniques used in search engines. As soon as an appropriate label LIB has been identified within the coding file, the classification codes Cl and C2 corresponding respectively to the two classification systems CLl, CL2 make it possible to activate dynamic links LD1, LD2 with functionalities available from from the doctor’s microcomputer. For example, the classification code C2 can automatically activate a dynamic link with a scientific database FS to provide scientific and medical information IS corresponding to the diagnosis object of the request and with a pharmaceutical database FM to provide IM information on reference pharmaceutical specialties, while the other classification code C1 provides a dynamic link with a FG management functionality to automatically supply this functionality with GI information specifically requiring a classification code corresponding to the first classification system. In addition, we can provide an external relationship via the DFCC code with table functions.

  FD countdown or FJ update functions.

  Furthermore, from an initial FC coding file offered at choice, the user can simply select items that meet his expectations, for example relating to a given patient, and thus constitute an FR file resulting from his choices. He can then use this result file FR resulting from operations for choosing CI items, which thus represents an extract from the initial file FC offered at choice, in the same

conditions as the latter.

  -9 It should be noted that the functionalities of the coding file generated with the method according to the invention

  are part of a relationship context, for example of relational databases type and implement5 the IT tools conventionally used in this context.

  The method according to the invention can also find an advantageous application in terms of transcoding between two coexisting classification systems, as illustrated in FIG. 3. Thus, from a code "abc" belonging to a first classification system CLl , it is possible to identify all of the items in the FC codification file to which this first "abc" code has been assigned. These identified items each have a unique DFCC code xyz, .., x'y'z ', .., x''y''z' 'and are also each assigned a classification code according to the second classification system. CL2. These classification codes ajy, .., a '' 'y', .. a constitute the result of the transcoding operation carried out on the initial abc code. This basic operation can be automated to provide full transcoding between the two classification systems. This transcoding is made possible thanks to the specific structure of the coding file. Furthermore, it can advantageously be provided that the use of the file for transcoding purposes is modulated as a function of frequency and origin characteristics of the items, accessible via the DFCC code, in particular to allow allocations.

  proportional in "l: n" type relationships.

  For the creation of a system for generating a coding file according to the invention, it is possible to use conventional means for entering, processing, storing and searching for information. The techniques used in the maintenance and management of relational databases can

  - 10 - naturally be called upon.

  Of course, the invention is not limited to the examples which have just been described and numerous modifications can be made to these examples without departing from the scope of the invention. In particular, the number of classifications used is not limited to two and

  other techniques for assigning the unique DFCC code may be considered. Furthermore, the present invention can find applications in other fields of information science and technology, in particular in electronic document management.

- 11-

Claims (14)

  1. Method for generating a codification file (FC) in an information space, in particular medical information, stored in at least one database and which can be classified according to a plurality of classification systems (CLl, CL2), characterized in that this coding file (FC) is generated by accumulation of items (LI) written in natural language10, and in that each item (LI) within said coding file (FC) is assigned several classification codes (Cl, C2) each belonging to a classification system (CL1, CL2) among the plurality of classification systems, so that the classification codes thus allocated provide links (LDl,   LD2) with other functionalities or databases.   2. Method according to claim 1, characterized in that each item within the coding file (FC), is also assigned a unique code (DFCC) constituting the smallest common part between this item and the different codes of classification assigned to said item. 3. Method according to claim 2, characterized in that the unique code (DFCC) attached to each item within the coding file is assigned sequentially.   4. Method according to one of claims 2 or 3,   characterized in that the unique code (DFCC) is arranged to provide a simple link with other characteristics of the item, in particular   characteristics of previous observations. - 12 -   5. Method according to claim 4, characterized in that a user can select from the coding file (FC) items meeting his expectations, so as to constitute a result file (FR) extracted from the coding file (FC ) and provided with the same link characteristics with bases and exterior features.   6. Method according to one of claims 2 to 5,   characterized in that it is further arranged to provide transcoding between a first classification system (CL1) and a second classification system (CL2) among the plurality of classification systems via the file   of codification (FC) generated with this process.   7. Method according to claim 6, characterized in that this transcoding takes into account the characteristics of the items, in particular frequency characteristics, accessible via the code single (DFCC).   8. Method according to one of claims   above, characterized in that it is arranged to generate, for each line of the coding file, a label (LIB) written in natural language and constituting the item, a unique code (DFCC) attached to this label, and several codes (Ci, C2) from several classifications (CL1, CL2).   9. System for generating a codification file (FC) in an information space, in particular medical information, stored in at least one database and which can be classified according to a plurality of classification systems (CL1, CL2), implementing the method according to any one of - 13 - previous claims,   characterized in that it comprises means for accumulating within this codification file (FC) items written in natural language by users of this database, and means for assigning to each item within said codification file (FC) several classification codes (Cl, C2) each belonging to a classification system (CL1, CL2) among the plurality of classification systems, so that the classification codes (Cl, C2) thus assigned provide links with other features or data base.   10. System according to claim 9, characterized in that it further comprises means for assigning to each item within the coding file (FC) a unique code (DFCC) constituting the smallest common part between this item and the different codes of   classification assigned to said item.   11. System according to one of claims 9 or 10,   characterized in that it further comprises means for transcoding between a first classification system (CL1) and a second classification system (CL2) among the plurality of classification systems via the coding file (FC) generated with this process.   12. Application of the method according to one of   claims 1 to 8 to a medical information system,   in which each item consists of a description of diseases, diagnoses or reasons for medical consultation. 13. Application of the method according to one of the   - 14 - claims 1 to 8 to an information system   medical, characterized in that the codification file in particular provides access to databases and functionalities which use or contain5 at least one of the classifications used in said codification file or in a file resulting from a   choice made within said coding file.