Related Experiment Videos
Using semantic distance for the efficient coding of medical concepts
C Bousquet1, M C Jaulent, G Chatellier
1Medical Informatics Department, Broussais Hospital, Paris, France.
Insights
This study introduces a semantic distance method to identify related medical concepts within controlled vocabularies, successfully distinguishing between ischemic and non-ischemic diseases.
Area of Science:
- Medical Informatics
- Computational Linguistics
Background:
- Controlled vocabularies like ICD-10 and SNOMED are crucial for medical data standardization.
- Accurate concept mapping between different terminologies is essential for interoperability.
Purpose of the Study:
- To develop and validate a method for identifying nearest neighbors of medical concepts using semantic distance.
- To assess the utility of semantic distance in differentiating between related and unrelated medical concepts.
Main Methods:
- Projected 392 cardiovascular concepts from ICD-10 onto SNOMED III axes.
- Calculated distances using Lp norm for concept proximity analysis.
- Validated the method by identifying nearest neighbors for ten ICD-10 cardiovascular diagnoses.
Main Results:
- Demonstrated a significant semantic distance (p < 0.0001) between sets of ischemic and non-ischemic diseases.
- Successfully validated the nearest neighbor identification for cardiovascular diagnoses.
- The method shows promise for improving medical concept mapping.
Conclusions:
- The semantic distance approach offers a novel way to navigate and relate medical concepts.
- Future integration with SNOMED-RT could enhance the method's precision.
- Further development is needed to create an ideal automated medical coding tool.
Objective:
To use the notion of semantic distance to find the nearest neighbors of a medical concept in a controlled vocabulary.
Material And Method:
392 concepts from the cardiovascular chapter of the ICD-10 were projected on the axes of SNOMED III. Distances were measured on each axis and the resulting distance was found using a Lp norm.
Results:
The distance between a set of ischemic diseases and a set of non-ischemic diseases was significant (p < 0.0001). Our method was validated by finding the k nearest neighbors of ten different diagnoses from the ICD-10 cardiovascular chapter.
Discussion:
The availability of SNOMED-RT should improve our method. Several more steps are necessary to provide an ideal coding tool.