Tooth vs. Teeth

Trying out the forum instead of e-mail...

I have been looking into the anatomy model after our discussion yesterday. Maybe you found the perfect solution, very sorry that I had to leave at schedule.

Just to clarify for my understanding:

In SNOMED plural is a child of singular, correct? This is contradictory to the rules Joel suggested - and maybe counter-intuitive to most. (At least to me). BUT looking deeper it makes sense. You go down the hierarchy to the point where you reach "the structure", i.e. tooth structure. It will have a child of tooth part, entire tooth, all the individual groups (molars, incisors etc) - and all teeth. All teeth will again have a child of complete dentition ("entire" all teeth).

If I understand this correct it means that a disorder will have a general description (demineralized tooth) with a child of affecting all teeth (demineralization of all teeth). Correct?

But the problem is then - what about a disorder affecting multiple instances of the same structure - but not all? (demineralization of multiple teeth)

Given the logic above it should be a child of the general disorder, and a sibling of affecting all teeth. But there is no anatomy concept for multiple but not all teeth.

Any thoughts?

If any of you could give me a few concepts I should look at, I don't remember them from yesterday, I can model them and see how it will look.

Jørn

A plural can be a child of a singular because:

• Multiple "Teeth" can be an instance of a singular "tooth", "teeth" IS-A "tooth" 
• "Tooth" cannot be a instance of "teeth", a single "tooth" is not a multiple "teeth", "tooth IS NOT A "teeth"

Jim has also suggested that we try and create a body structure of "Multiple teeth (body structure)" as a sibling of "Tooth structure" to take care of the situation that Jorn describes here.

If anyone can name three or four concept for "teeth" within clinical findings I can make a task in authoring trying out a "Multiple teeth (body structure)", and name that as a finding site for teeth concepts.

Then we can see how this would play out in the hierarchy. And then set the "rules"

Jorn

I'm not sure if this is what you are looking for Jorn:

Concept ID 109734002 Dental restoration failure of anatomical integrity regarding adjacent teeth Attribute value 109731005 Tooth restoration failure (finding)

Concept ID 734009000 Diastema of teeth (finding)  Attribute value 38199008 Tooth structure (body structure)

Concept ID 734025005 Excessive spacing of fully erupted teeth (finding)  Attribute value 38199008 Tooth structure (body structure)

Concept ID 366228003 Finding of interincisal teeth angle (finding)  Attribute value 366223007 Finding of tooth angulation (finding)

A plural can be a child of a singular.

Can a singular be a child of a singular (rephrasing Joel's definition during our call):

• "Tooth" can be an instance of a singular "tooth"
• Concept ID 278679009 Tooth discolored (finding) Attribute value 366229006 Finding of tooth color (finding)

Further,

Can a plural be a child of a plural (again rephrasing Joel)

• "Teeth" can be an instance of a multiple "Teeth"
• Concept ID 163156007 On examination - discolored teeth (finding)  Attribute value 270479002 On examination - teeth examined (finding)

Thank you Roger,

these are just the kind of examples I'm looking for. I will try to model them within one task and we can view them in the browser.

When it comes to the "rules" I think you are right.

• a singular can be a child of singular
• a plural can be a child of plural

What is opposite of what Joel proposed is that plural is a child of singular. (Not singular as a child of plural as I understood Joel). But how this will look in the dental world remains to be seen. So far that would be our third "rule".

Singular as a child of plural would then be out.

Here are the four, actually six, rules as, I believe, Joel proposed:

1)  Tooth is singular, teeth is plural

2) If singular in concept, body structure needs to be singular  (within spreadsheets developed by Penni)

3) If plural in concept, and attribute is plural, it is OK

4) If it is singular in concept, the attribute cannot be plural

5) If it is plural in concept, it is possible that singular attribute is OK (this appears to be the case from Penni's and Jorn's discussions above)

6)  If singular in concept, body structure attribute cannot be plural  (this is similar to 2, if we consider 2 to be an absolute)

We spent considerable time discussing #1 and everyone agreed that this is the lens we should use.  We did not spend any time discussing variations of attributes- whether they are "is a" relationships that are hierarchical (and therefore are parent/child and sibling relationships or other types of relationships), as we look at this.  Do we need to define the rules differently, understand whether they might impact other attributes differently using this lens, or are we basically going to look at the "is a" relationships?

I will be anxious to see if Jorn's modeling is consistent with the rules.  If so, are we ready to review the entire block of terms,  using these rules, to determine where changes are necessary?  If the modeleing does hold, then I think we should add the multiple teeth concept per Jim's suggestion.

Hi all,

And a very happy new year to all of you reading this discussion.

I have now done a minor modelling in task dentistry-112, adding multiple teeth (body structure) and changing attributes for a few concepts. In addition to the ones suggested by Roger, I have changed attributes for Teeth covered in plaque (finding) SCTID: 276453000

On "Concept ID 109734002 Dental restoration failure of anatomical integrity regarding adjacent teeth Attribute value 109731005 Tooth restoration failure (finding)" I'm a bit more uncertain, maybe misinterpreting the meaning of this.

The way I see it "dental restoration" in this context is a singular restoration. It is failing due to the relationship it is establishing to adjacent teeth. So it makes sense that it "is a" tooth restoration failure" (singular). I could be claimed that it has a finding site of multiple teeth, or rather a finding site of interdental relationship between multiple teeth.

Anyhow. Here I just added the finding site of multiple teeth.

So, please take a look and see if it holds up so far. If it does there is a bit more to be done before we can conclude that the rules are OK.

Jorn

Hi All,

Many of you will be able to look at Jorn's task via the TS Browser (https://authoring.ihtsdotools.org/browser/) using the login: snowowl/snowowl, but remember to go to the "Dentistry" project and Jorn's task # Dentistry-112 (see the drop down boxes at the top of the screen)

To me there seems 

I had a question Jorn, about the example of 276453000|Teeth covered in plaque (finding)| is this ever normal? or should it be a disorder?  If it is a disorder it could be fully defined by role grouping your new finding site concept with 1522000|Plaque (morphologic abnormality)|

That is almost a philosophical question. What is normal? By nature teeth are covered by pellicle. Plaque are deposits of bacteria adhering to the pellicle, eventually being calcified by salivary minerals - and ending ut as calculus.

The thing is that most people have it, at least to some mild extent in the afternoon. But the more comprehensive coverage of multiple teeth over time would not be normal. It may indicate insuficient brushing or lack of mucular funktion - among other reasons. But it is not itself a disorder. It may cause different disorders, periodontitis, caries etc.

I am struggeling with this one. I would like us to fully define it, but it comes down to the initial question - what is normal? What is a disorder? Could we take a look together on Zoom? Just see what happens if we define it one way or the other. I'm thinking it at least should have a causative agent of bacteria.

Looking at this I was wondering if we should also make a grouper "disorder of multiple teeth" (tooth disorders affecting more than one tooth at the same time). Or maybe find another grouper containing findings related to the dental arch, but not the tooth itself.

You wrote "To me there seems..." Is there something missing?

Jørn

Teeth covered in plaque may, in fact, be the normal state.    I would struggle to call it a disorder, even though we, in dentistry, do not consider it to be the best state and there is certainly evidence that the bacteria in plaque leads to disorders such as periodontal disease or at least gingivitis when it is in contact with the gum tissue for some period of time (varies with individuals).

I think this is better off as a finding or even an observation?

I don't think normal is equal to healthy. I prefer disorder for this, but I am amenable to finding. I disagree with observation.

Dave makes a good point and I should have used the word common or most common state rather than normal.  There are likely many "common" conditions in lesser developed countries(and elsewhere) that should be traceable in some way and contribute to or relate to disorders/diseases.  I am struggling to see how we would use this information to do anything but document a very common condition. If the overwhelming majority of the population has plaque on their teeth, I cannot see how we could develop any decision support or even correlations that would let us understand health better.

The remote, but possible, utility of an observable is that we might be able to (eventually) develop a scale to create smaller sets of the very large population and see if there is a difference in relationships to disorders.  However, that is not my choice of approach either.