Research & Innovation

Publications

Panacea, a semantic-enabled drug recommendations discovery framework

J Biomed Semantics. 2014 5(13). doi: 10.1186/2041-1480-5-13

Doulaverakis C, Nikolaidis G, Kleontas A, Kompatsiaris I.
Source: Centre for Research and Technology Hellas, Information Technologies Institute, Thessaloniki, Greece. doulaver@iti.gr.

BACKGROUND: Personalized drug prescription can be benefited from the use of intelligent information management and sharing. International standard classifications and terminologies have been developed in order to provide unique and unambiguous information representation. Such standards can be used as the basis of automated decision support systems for providing drug-drug and drug-disease interaction discovery. Additionally, Semantic Web technologies have been proposed in earlier works, in order to support such systems.

RESULTS: The paper presents Panacea, a semantic framework capable of offering drug-drug and drug-diseases interaction discovery. For enabling this kind of service, medical information and terminology had to be translated to ontological terms and be appropriately coupled with medical knowledge of the field. International standard classifications and terminologies, provide the backbone of the common representation of medical data while the medical knowledge of drug interactions is represented by a rule base which makes use of the aforementioned standards. Representation is based on a lightweight ontology. A layered reasoning approach is implemented where at the first layer ontological inference is used in order to discover underlying knowledge, while at the second layer a two-step rule selection strategy is followed resulting in a computationally efficient reasoning approach. Details of the system architecture are presented while also giving an outline of the difficulties that had to be overcome.

CONCLUSIONS: Panacea is evaluated both in terms of quality of recommendations against real clinical data and performance. The quality recommendation gave useful insights regarding requirements for real world deployment and revealed several parameters that affected the recommendation results. Performance-wise, Panacea is compared to a previous published work by the authors, a service for drug recommendations named GalenOWL, and presents their differences in modeling and approach to the problem, while also pinpointing the advantages of Panacea. Overall, the paper presents a framework for providing an efficient drug recommendations service where Semantic Web technologies are coupled with traditional business rule engines.

GalenOWL: Ontology-based drug recommendations discovery

J Biomed Semantics. 2012 Dec 20;3(1):14. doi: 10.1186/2041-1480-3-14

Doulaverakis C, Nikolaidis G, Kleontas A, Kompatsiaris I.
Source: Centre for Research and Technology Hellas, Information Technologies Institute, Thessaloniki, Greece. doulaver@iti.gr.

BACKGROUND: Identification of drug-drug and drug-diseases interactions can pose a difficult problem to cope with, as the increasingly large number of available drugs coupled with the ongoing research activities in the pharmaceutical domain, make the task of discovering relevant information difficult. Although international standards, such as the ICD-10 classification and the UNII registration, have been developed in order to enable efficient knowledge sharing, medical staff needs to be constantly updated in order to effectively discover drug interactions before prescription. The use of Semantic Web technologies has been proposed in earlier works, in order to tackle this problem.

RESULTS: This work presents a semantic-enabled online service, named GalenOWL, capable of offering real time drug-drug and drug-diseases interaction discovery. For enabling this kind of service, medical information and terminology had to be translated to ontological terms and be appropriately coupled with medical knowledge of the field. International standards such as the aforementioned ICD-10 and UNII, provide the backbone of the common representation of medical data, while the medical knowledge of drug interactions is represented by a rule base which makes use of the aforementioned standards. Details of the system architecture are presented while also giving an outline of the difficulties that had to be overcome. A comparison of the developed ontology-based system with a similar system developed using a traditional business logic rule engine is performed, giving insights on the advantages and drawbacks of both implementations.

CONCLUSIONS: The use of Semantic Web technologies has been found to be a good match for developing drug recommendation systems. Ontologies can effectively encapsulate medical knowledge and rule-based reasoning can capture and encode the drug interactions knowledge.