SIMI 2012 : Semantic Interoperability in Medical Informatics
Dates:
When May 27, 2012 – May 27, 2012
Where Heraklion (Crete), Greece
Submission Deadline Mar 4, 2012
Notification Due Apr 1, 2012
Final Version Due Apr 15, 2012
From the call for papers:
To gather data on potential application to new diseases and disorders is increasingly to be not only a means for evaluating the effectiveness of new medicine and pharmaceutical formulas but also for experimenting on existing drugs and their appliance to new diseases and disorders. Although the wealth of published non-clinical and clinical information is increasing rapidly, the overall number of new active substances undergoing regulatory review is gradually falling, whereas pharmaceutical companies tend to prefer launching modified versions of existing drugs, which present reduced risk of failure and can generate generous profits. In the meanwhile, market numbers depict the great difficulty faced by clinical trials in successfully translating basic research into effective therapies for the patients. In fact, success rates, from first dose in man in clinical trials to registration of the drug and release in the market, are only about 11% across indications. But, even if a treatment reaches the broad patient population through healthcare, it may prove not to be as effective and/or safe as indicated in the clinical research findings.
Within this context, bridging basic science to clinical practice comprises a new scientific challenge which can result in successful clinical applications with low financial cost. The efficacy of clinical trials, in combination with the mitigation of patients’ health risks, requires the pursuit of a number of aspects that need to be addressed ranging from the aggregation of data from various heterogeneous distributed sources (such as electronic health records – EHRs, disease and drug data sources, etc) to the intelligent processing of this data based on the study-specific requirements for choosing the “right” target population for the therapy and in the end selecting the patients eligible for recruitment.
Data collection poses a significant challenge for investigators, due to the non-interoperable heterogeneous distributed data sources involved in the life sciences domain. A great amount of medical information crucial to the success of a clinical trial could be hidden inside a variety of information systems that do not share the same semantics and/or structure or adhere to widely deployed clinical data standards. Especially in the case of EHRs, the wealth of information within them, which could provide important information and allow of knowledge enrichment in the clinical trial domain (during test of hypothesis generation and study design) as well as act as a fast and reliable bridge between study requirements for recruitment and patients who would like to participate in them, still remains unlinked from the clinical trial lifecycle posing restrictions in the overall process. In addition, methods for efficient literature search and hypothesis validation are needed, so that principal investigators can research efficiently on new clinical trial cases.
The goal of the proposed workshop is to foster exchange of ideas and offer a suitable forum for discussions among researchers and developers on great challenges that are posed in the effort of combining information underlying the large number of heterogeneous data sources and knowledge bases in life sciences, including: – Strong multi-level (semantic, structural, syntactic, interface) heterogeneity issues in clinical research and healthcare domains – Semantic interoperability both at schema and data/instance level – Handling of unstructured information, i.e., literature articles – Reasoning on the wealth of existing data (published findings, background knowledge on diseases, drugs, targets, Electronic Health Records) can boost and enhance clinical research and clinical care processes – Acquisition/extraction of new knowledge from published information and Electronic Health Records – Enhanced matching between clinicians as well as patients΅¦ needs and available informational content
Apologies for the length of the quote but this is a tough nut that simply saying “topic maps,” isn’t going to solve. As described above, there is a set of domains, each with its own information gathering, processing and storage practices, none of which are going to change rapidly, or consistently.
Although I think topic maps can play a role in solving this sort of issue, it will be by being the “integration rain drop” that starts with some obvious integration issue and solves it and only it. Does not try to be a solution for every issue or requirement. Having solved one, it then spreads out to solve another one.
The key is going to be the delivery of clear and practical advantages in concrete situations.
One approach could be to identify current semantic integration efforts (which tend to have global aspirations) and effect semantic mappings between those solutions. Which has the advantage of allowing the advocates of those systems to continue while a topic map can offer other systems an integration of data from those parts.