Table 2 Ontologies related to BTO that model MS aspects. Usage TH indicates that the ontology is supposed to be treated as thesaurus, KB indicates that the ontology is used as the schema for a knowledge base; AI indicates that the ontology is used as additional information for machine learning approaches
Name | AEDSS Application Ontology | |
Usage | AI | |
Online | No | |
Description | Ontology used to determine automatically the EDSS (AEDSS), using an expert system. | |
Differences | Focuses only on EDSS, no other elements of MS modeled. | |
Name | — | |
Usage | AI, KB | |
KB Description | 240 magnetic resonance images representing white matter lesions, annotated according to the ontology. | |
Online | No | |
Description | Ontology used to annotate white matter lesions images. Such annotations are further used in a machine-learning algorithm to classify images. | |
Differences | Focuses only on white matter lesions and images. | |
[18] | Name | Multiple Sclerosis Patient Data Ontology (MSPD) |
Usage | KB | |
Online | Partially (Released: 2014 – Last Update: 2016). The ontology is available at https://github.com/mark-jensen/mspd/blob/master/mspd_06.owl but the URIs for classes defined within MSPD are broken or non-existent. | |
KB Description | Data concerning 10,000 patients and 17,000 follow-up visits from the NYSMSC (not available publicly) | |
Description | Extension of NDO [8,9,10], used to encode self-assessment of the disability perceived by the patients. Such self-assessments are then used to determine the prevalence of different characteristics among the population’s subgroups. | |
Differences | MSPD does not consider aspects related to clinical events, besides a set of concepts used to diagnose the MS. BTO, on the other hand, besides diagnostic assays for MS, includes also clinical history details (e.g., previous surgeries, traumas, pregnancies, genetic data). |