Mapping and matching concepts: Difference between revisions

From Endeavour Knowledge Base
No edit summary
 
(5 intermediate revisions by the same user not shown)
Line 1: Line 1:
== Mapping API ==
== Mapping approach ==
The information model mapping API is used, when given contextual data, to return both a property concept (representing the context) and a value concept (representing the value within that context). The structure of the mapping module allows multiple contexts to reference a common mapping node (for example when the same type of value is received from multiple sources) and also multiple value types to be referenced from a common mapping node.
From an ontological perspective it could be said that 'A' in the context of 'C', when manipulated by some function, is probably equivalent to 'B'. In OWL this would be denoted as an equivalent class axiom.


The logic of the version 1 live mapping is described  more fully in the [[Information Model Mapping (v1)|model mapping article]].
From a slightly more process based perspective it could be said that 'A' with context 'C' -> Maps to (via a function) > 'B'


== Managing codes and taxonomies ==
A and B may be entities, properties, objects or literals. A and B may be collections such that there is no one to one map.
Information consists of ideas. Another word for an idea is a 'concept' . A concept may be named,( in which case the meaning of the concept can usually be understood), or they may be an unnamed expression, which is made up of a set of interrelated named or unnamed concepts.  


For example the term "chest pain" implies the idea of a pain in the chest. In Snomed-CT it is a named concept. "Chest pain, worsened by exercise" may be an example of an expression style concept made up from the concept of "chest pain", and the statement that it is "made worse by -> exercise". In this case “made worse by” and “exercise” are both different concepts but no author has yet created a single named concept for this expression.
When transforming data from one form (the source) to another form (the target), a process of transformation is undertaken. That transformation uses software.


The new generation of health record management systems tend towards the recording of concepts, with the objective being for the record entry to closely match the idea behind the entry. These types of concepts can be called term based concepts as the term is the thing that describes the idea.
There are general two broad approaches to transformation:


A modern term based concept is defined in relation to other concepts by a set of assertions indicating whether the concept is equivalent to, or a subtype of, a set of other concepts. The standard approach to this is via the use of Description Logic (DL). By using DL, a computer can automatically classify a concept which can result in a computer deducing additional knowledge over and above the human who created the concept. Snomed-CT is the worlds largest ontology of healthcare term based concepts and is authored using DL. A collection of concepts defined in this way constitute an "Ontology" and there is a standard language OWL that is used to represent the definitions.
a) Each transformation from A to B from each context C is written, or at least generated, in source code, which is executed by some form of transformation engine.


These types of concepts are referred to as "Core concepts"
b) A transformation engine includes software that uses a 'Transform Map' which is a separate resource, describing the way in which A in context C, maps to B.
 
As one would expect, in an information model, the latter approach is used, with the transform maps having a visualisation capability for humans to debate, decide, and assure, that the equivalence exists to the extent that is safe.
 
Thus, the role of the information model is to provide a library of maps using, a meta model of classes that define the maps.
 
In line with all of the information model, these classes are defined using the same information model language that the rest of the model uses, with a set of meta classes, conceptually equivalent to transformation languages.
 
More specifically, the mapping meta classes have been informed by FHIR Mapping language, the RML semantic web based mapping language, and the simpler concept maps provided by many sources.
 
Types of mapping can be broadly categorised into 3 categories
 
# Things which map entities and properties from one type to another
# Things which map codes or concepts or text.
# Things which map property values using the source properties and values as key to a look up to a reference entity.
 
<br />
 
== Mapping entities and properties ==
This includes:
 
a) The provision of definitions that determine how one set of source entities (or resources, messages or tables) map to a target entity.
 
b) the provision of a reference transform engine that illustrates how these maps can be used in the real world.
<br />
== Mapping codes and taxonomies ==
The new generation of health record management systems tend towards the recording of concepts, with the objective being for the record entry to closely match the idea behind the entry. These types of concepts can be called term based concepts as the term is the thing that describes the idea.
 
A modern term based concept is defined in relation to other concepts by a set of assertions indicating whether the concept is equivalent to, or a subtype of, a set of other concepts. This is normally referred to as an ontology.
 
Snomed-CT is the worlds largest ontology of healthcare term based concepts and is authored using a form of Description logic, which enables a reasoner to automatically classify a concept according to its properties.


The idea of codes originated from a different starting point. The intention of a coded entry is to ''pre-classify'' an entry before it is recorded. The code is designed for a particular set of business processes e.g. analytics or payment and it is important to understand the context in which a code has been used.  A coded concept, being pre-classified, relies on categorisation of the codes, and that classification may or may not imply that one code is a subtype of another. Nothing can be inferred from a code other than its relation to another code as authored.  Consequently, as the philosophy is different, code based concepts have to be dealt with differently from term based concepts, even if they seem to saying the same thing.
The idea of codes originated from a different starting point. The intention of a coded entry is to ''pre-classify'' an entry before it is recorded. The code is designed for a particular set of business processes e.g. analytics or payment and it is important to understand the context in which a code has been used.  A coded concept, being pre-classified, relies on categorisation of the codes, and that classification may or may not imply that one code is a subtype of another. Nothing can be inferred from a code other than its relation to another code as authored.  Consequently, as the philosophy is different, code based concepts have to be dealt with differently from term based concepts, even if they seem to saying the same thing.
Line 19: Line 48:
Because of their history, it is not always possible to assert the exact meaning of a code. However, it is often the case that meaning can be inferred or approximated from a coded entry. With  preference to move to an ontology, this inference can be achieved via the use of a mapping process that matches  coded concepts to term based concepts that are identified from a code.
Because of their history, it is not always possible to assert the exact meaning of a code. However, it is often the case that meaning can be inferred or approximated from a coded entry. With  preference to move to an ontology, this inference can be achieved via the use of a mapping process that matches  coded concepts to term based concepts that are identified from a code.


These types of concepts are referred to as "legacy concepts"
These types of concepts are referred to as "legacy concepts , or non core concepts"


There are two strategies to link legacy code concepts to core concepts.
There are two strategies to link none core code concepts to core concepts.


1. A coded term may be stated confidently to be the same as, or a variation on, a concept. Typically code systems like Read2 or CTV3 can be dealt with in this way because they are designed to try and capture the idea in the clinicians mind, and they have been incorporated as concepts anyway. Likewise many system supplier codes have been created in this way. In this case the term code can be said to be a term code of the concept. Read2 G33 - Angina pectoris is a term code for the concept of angina pectoris.
1. A coded term may be stated confidently to be the same as, or a variation on, a concept. Typically code systems like Read2 or CTV3 can be dealt with in this way because they are designed to try and capture the idea in the clinicians mind, and they have been incorporated as concepts anyway. Likewise many system supplier codes have been created in this way. In this case the term code can be said to be a term code of the concept. Read2 G33 - Angina pectoris is a term code for the concept of angina pectoris.
Line 61: Line 90:
                       [
                       [
                           :oneOf  icd10:O24.3 icd10:O24.9 O24.4]
                           :oneOf  icd10:O24.3 icd10:O24.9 O24.4]
</pre><br />
</pre><br /><br />
== Mapping  source fields ==
It is equally common to need to map source fields to core data model properties. In the information model a property is considered just another concept.
 
However, to use mappings for source fields it is necessary to use the context in which the source fields exist.
 
In the above examples, coded concepts were considered as context independent in the sense that the same code used by many providers and many systems would generally mean the same thing and can be treated the same way.
 
 
In the same way that codes can be mapped, so can source resource types such as tables or fields, message types or message segments. Mapping may involve functional transformation
 
=== Defining source context ===
The first step in managing source concepts is to define the concept in the context of the originator of the data. This employs the use of a context object usually sent as parameters through the REST API. For example the following<pre>
 
{
  "organisation": "Barts",
  "system" : "CernerMillenium",
  "message" :"cds_type_130",
  "field" :"admin_cat_code"
}
 
</pre>
 
=== Mapping nodes ===
A second step is for the mapping author  to identify whether the source context  is equivalent to another source contexts. This is done in order to rationalise the number of mappings steps needed between a source concept and the final target concept.
=== Matching to concept ===
The third step involves creation of a source concept and a core concept. In the above example the source is concept has been mapped to a core concept and the core concept is returned<pre>
bc:BC_xyz
  rdfs:label "admin_cat_code"
 
and potential core map
 
im:administrative_category_code_on_admission
          :matchedTo
          bc:BC_xysdasdasd.
 
 
</pre>The information model has fully defined the administrative category property as a property of a subclass of encounter dealing with hospital stays. Consequently the source system's table and field can be fully mapped to the common model field.
<br />
== Legacy codes, terms and term codes==
Legacy or  local codes also require context and the same approach is used as described above. In this case a local code 12345655 mat have different meaning in a ddifferent system. A local code may or may not have any sort of code scheme (the scheme then being implied by the context). In the following example, Barts trust has its own local cerner code scheme.<pre>
{
  "organisation": "Barts",
  "system" : "CernerMillenium",
  "message" :"cds_type_130",
  "field" :"admin_cat_code",
  "codeScheme" :"BartsLocalCodes",
  "code" :123445556
}
 
</pre>
 
== Enumerated values ==
The NHS Data dictionary employs enumerated numeric values for many of its field contents.
 
These are treated as local codes using the context to create a field specific code scheme. In many cases local trusts have extended the national scheme (or even changed the scheme), and as a result mapping nodes are used for the common codes.
 
==Decision process for handling codes and terms==
When incorporating codes and terms into the ontology, there are 4 categories of approach, the selected approach being dependent on the semantic relationship between the legacy and core concept. The categories are:
 
#Creation of a local concept and creation of a simple "matched to" map between a core concept and a legacy concept. This implies that the legacy concept is a subtype of, or equivalent to the core concept. Examples of these are supplier local codes, Read 2, TPP
#Creation of a local or legacy concept and a complex "combination of" + some of or one of "mapped to" between a core concept and a legacy concept. This implies a more nuanced relationship meaning that the user may elect how to use the maps when querying records.
#Mapping of the local context object to a mapping node which maps to a core concept i.e. the core concept is used throughout the data store and the local context is replaced.
#Mapping of the local context object to a mapping node which maps to a legacy concept which may or may not map from a core concept
 
<br />

Latest revision as of 10:53, 26 October 2022

Mapping approach

From an ontological perspective it could be said that 'A' in the context of 'C', when manipulated by some function, is probably equivalent to 'B'. In OWL this would be denoted as an equivalent class axiom.

From a slightly more process based perspective it could be said that 'A' with context 'C' -> Maps to (via a function) > 'B'

A and B may be entities, properties, objects or literals. A and B may be collections such that there is no one to one map.

When transforming data from one form (the source) to another form (the target), a process of transformation is undertaken. That transformation uses software.

There are general two broad approaches to transformation:

a) Each transformation from A to B from each context C is written, or at least generated, in source code, which is executed by some form of transformation engine.

b) A transformation engine includes software that uses a 'Transform Map' which is a separate resource, describing the way in which A in context C, maps to B.

As one would expect, in an information model, the latter approach is used, with the transform maps having a visualisation capability for humans to debate, decide, and assure, that the equivalence exists to the extent that is safe.

Thus, the role of the information model is to provide a library of maps using, a meta model of classes that define the maps.

In line with all of the information model, these classes are defined using the same information model language that the rest of the model uses, with a set of meta classes, conceptually equivalent to transformation languages.

More specifically, the mapping meta classes have been informed by FHIR Mapping language, the RML semantic web based mapping language, and the simpler concept maps provided by many sources.

Types of mapping can be broadly categorised into 3 categories

  1. Things which map entities and properties from one type to another
  2. Things which map codes or concepts or text.
  3. Things which map property values using the source properties and values as key to a look up to a reference entity.


Mapping entities and properties

This includes:

a) The provision of definitions that determine how one set of source entities (or resources, messages or tables) map to a target entity.

b) the provision of a reference transform engine that illustrates how these maps can be used in the real world.

Mapping codes and taxonomies

The new generation of health record management systems tend towards the recording of concepts, with the objective being for the record entry to closely match the idea behind the entry. These types of concepts can be called term based concepts as the term is the thing that describes the idea.

A modern term based concept is defined in relation to other concepts by a set of assertions indicating whether the concept is equivalent to, or a subtype of, a set of other concepts. This is normally referred to as an ontology.

Snomed-CT is the worlds largest ontology of healthcare term based concepts and is authored using a form of Description logic, which enables a reasoner to automatically classify a concept according to its properties.

The idea of codes originated from a different starting point. The intention of a coded entry is to pre-classify an entry before it is recorded. The code is designed for a particular set of business processes e.g. analytics or payment and it is important to understand the context in which a code has been used. A coded concept, being pre-classified, relies on categorisation of the codes, and that classification may or may not imply that one code is a subtype of another. Nothing can be inferred from a code other than its relation to another code as authored. Consequently, as the philosophy is different, code based concepts have to be dealt with differently from term based concepts, even if they seem to saying the same thing.

Because of their history, it is not always possible to assert the exact meaning of a code. However, it is often the case that meaning can be inferred or approximated from a coded entry. With preference to move to an ontology, this inference can be achieved via the use of a mapping process that matches coded concepts to term based concepts that are identified from a code.

These types of concepts are referred to as "legacy concepts , or non core concepts"

There are two strategies to link none core code concepts to core concepts.

1. A coded term may be stated confidently to be the same as, or a variation on, a concept. Typically code systems like Read2 or CTV3 can be dealt with in this way because they are designed to try and capture the idea in the clinicians mind, and they have been incorporated as concepts anyway. Likewise many system supplier codes have been created in this way. In this case the term code can be said to be a term code of the concept. Read2 G33 - Angina pectoris is a term code for the concept of angina pectoris.

2. A coded term might be the same term as a concept but may have been entered without the assertion that is a true representation of a state. Typically code systems such as ICD10 and OPCS fall into this category. E11 - Diabetes type 2, seems to be the same as the concept of diabetes type 2, but was entered without clinician attestation and may have been approximated for payment purposes. In this case a legacy concept is produced and a map between this concept and the similar clinical concept is generated.

A map is just another form of relationship, but unlike an ontological equivalent or subclass axiom it implies that the relationship is an approximation. It is a sort of statement that something is possibly or probably similar to something else and thus has much less weight than an asserted relationship.

Legacy Code based concepts can be mapped to Core concepts , and this enables the use of the vast volumes of data already recorded in systems. Maps must be used with care as it is almost always the case that the use of a mapped code in a query is dependent on the purpose of the query. This means that mappings are more of a guide to the things to include rather than a confident statement of meaning. When querying records the query author may need to determine which codes to include or exclude on a case by case basis.

Maps between core concepts and legacy concepts

As mentioned above the relationships are managed as mappings which state the type or degree of match.

Maps generally fall into 2 patterns. These are illustrated in the context of code based concepts as follows:

Simple match

A core concept may be matched to many code based concepts. In a simple match the legacy concept is deemed to be probably equivalent to, or a subclass of. the code concept


sn:194828000 |Angina (disorder)
    :matchedTo emis:G33 |Angina Pectoris|.

Complex optional match

A concept may be matched to a number of alternative concepts and it is expected that a query author may wish to select these.

In this example, the concept : "Ketoacidotic coma due to diabetes mellitus (disorder)" has a complex map which is selection of either

a) Coma unspecified

and

b) one of either Diabetes mellitus in pregnancy: Pre-existing diabetes mellitus, unspecified, or Diabetes mellitus in pregnancy, unspecified, or Diabetes mellitus arising in pregnancy

In effect meaning that the compound entry in the record would need to have 2 icd 10 codes to fulfill the criteria. 

sn:26298008
  :hasMap [
       :combinationOf  [ 
                           :oneOf  icd10:R402 ] 
                       [
                           :oneOf  icd10:O24.3 icd10:O24.9 O24.4]



Retrieved from ‘https://wiki.endeavourhealth.org/index.php?title=Mapping_and_matching_concepts&oldid=6604’
Last modified
This page was last modified on 26 October 2022, at 10:53.