The Knowledge Model is pivotal for Decision Support Systems since it must allow for the representation of patient data information, decisional criteria sharing a common data model and enabling reasoning at different levels of abstraction.
The underlying data model is based on the generic Entity-Attribute-Value model, providing the conceptual skeleton for recording and structuring an ontology. The result is the Breast Cancer Knowledge Model (BCKM) under OWL containing i) knowledge about the Breast Cancer domain (procedures, examination results, tests results classification, etc..) and ii) knowledge about the patient case (age, pathological characteristics, etc).
The model for representing clinical inference knowledge relies on a generic production rule model consistent with the underlying data model in DESIREE Guideline Definition Language (DGDL). The DGDL uses the concepts in the BCKM in the decision rules allowing to benefit from the conceptual organisation of the BCKM (subsumption, complex concepts definitions, etc.) to mix both classification capabilities and rule-based reasoning.
CancerAddressing target audiences and expressing needs
- To raise awareness and possibly influence policy
To further disseminate the Breast Cancer Knowledge Model Ontology to be used and further refined in the framework of other R&I projects. To that end, we are open to openly disseminate and make available the Deliverable/Report describing the Ontology.
This is a fairly comprehensive knowledge model of breast cancer. So this provides the ground for further work. This knowledge model can provide the basis for building a set of axioms that explain the guideline knowledge base. This is an example of explainable AI.
- EU and Member State Policy-makers
- Research and Technology Organisations
R&D, Technology and Innovation aspects
The generation of new rules for the Breast Cancer Knowledge Model (BCKM) can be easily extrapolated to other domains. The representation of patient data information, decisional criteria sharing a common data model and the reasoning at different levels of abstraction can easily be adjusted to match the respective guidelines in other domains. Furthermore, the platform as such provides a comprehensive support to the clinical needs in the cancer treatment. The system as it is can be exactly deployed across hospitals and healthcare organisations following the same approach to that of the rules applied in the original facilities, as the rules and guidelines are international and cross-border, so the replicability can easily be followed.
Breast Cancer Knowledge Model (BCKM) is considered an emerging market without an industry leader and ample market potential, the development is certainly a scalable innovation within individual countries as well as on an international scale. Furthermore, the deployment of the platform can be easily integrated within the daily practice of Breast Units in hospital, which enhances the scalability of the system across clinical organisations, regardless of the country and requiring minimum dedicated training to the system by practitioners.
- Global
- Europe
