MIDAS is a collaborative project between Ohio Center of Excellence in Knowledge-enabled Computing (Kno.e.sis) and Milcord LLC. MIDAS is an effort to bring about the new era in building the Department of Defense Electronic Medical Record System under the Defense Healthcare Management System Modernization (DHMSM) effort.
MIDAS is a two-phase project. In the first phase, Kno.e.sis will demonstrate its healthcare related semantic technology in the cardiology domain, and in the second phase, Kno.e.sis will apply this technology to the musculoskeletal disease domain. Both these domains can have high impact on the health and well-being of US population and consequently on the US healthcare budget. Heart disease is the leading cause of death for both men and women in US. US reports 610, 000 deaths due to heart diseases annually and it is the cause for 25% of the deaths.<ref>http://www.cdc.gov/heartdisease/facts.htm</ref> Furthermore, it has been reported that each year 735,000 Americans have heart attack and 525, 000 of them were reported as first time attacks. Musculoskeletal disease is among the most disabling and costly condition suffered by the Americans. A worldwide study conducted in 2012 reported that 1.7 billion people in the world suffer from musculoskeletal diseases such as arthritis and back pain. It is recognized as the second greatest cause of disability, and has the 4th greatest impact on the overall health of the world population when considering both death and disability. In US, it is reported that more than 50% of adults suffer from musculoskeletal disease. This number escalates for people over the age of 65, where it has been reported that nearly 75% suffer from musculoskeletal disease in that population segment. This problem becomes even more significant with the aging population in US. By 2040, the population over 65 is projected to grow from 15% to 21%.<ref>http://www.boneandjointburden.org</ref>
Ultimately, the goal of the MIDAS project is to enhance the current capabilities of the Electronic Medical Record System used by the Department of Defense focusing on domains such as these. The broader agenda includes enhancing the communication between the physicians and the patients who are currently limited to the regular visits, extracting and exploiting the knowledge buried in the medical records to identify more effective treatment plans, and empowering the patients by enabling them to monitor their health status.
Specific Project Goals:
- To develop a prototype to recognize and annotate medical concepts (such as disorders, symptoms, procedures, and treatments) in free-text documents and map them to UMLS.
- To develop and demonstrate data-driven techniques to enhance and fill gaps in existing knowledge base or domain ontology using information in EMRs. This will be demonstrated by enriching an ontology using new statistical correlations surfaced by analyzing EMRs and then verified by domain expert. The ontology can be constituted by a variety of binary associations such as disorder-symptom, disorder-procedure, and disorder-treatment.
- Given structured information about patients, research and define metrics to compare them. Specifically, given structured information extracted from EMR records about the symptoms, diagnosis, procedures, and demographic information about patients, research approaches to compare two records, or rank order different records on the basis of similarity to a record, with an eye to exploit the treatment information present in the individual records or synthesized from a group of records for case-based reasoning.
- Building on the above foundation, develop architectural design and functional requirements for more comprehensive research prototype.
- Sujan Perera, Cory Henson, Krishnaprasad Thirunarayan, Amit Sheth, Suhas Nair, Data driven knowledge acquisition method for domain knowledge enrichment in the healthcare, IEEE International Conference Bioinformatics and Biomedicine 2012 (BIBM '12), pp.1-8, 4-7 Oct. 2012
- Sujan Perera, Amit Sheth, Krishnaprasad Thirunarayan, Suhas Nair and Neil Shah, Challenges in Understanding Clinical Notes: Why NLP Engines Fall Short and Where Background Knowledge Can Help, International Workshop on Data management & Analytics for healthcaRE (DARE) at ACM Conference of Information and Knowledge Management (CIKM), pp. 21-26, Burlingame, USA, Nov 1, 2013
- Sujan Perera, Cory Henson, Krishnaprasad Thirunarayan, Amit Sheth, Suhas Nair,Semantics Driven Approach for Knowledge Acquisition From EMRs, IEEE Journal of Biomedical and Health Informatics, vol.18, no.2, pp.515-524, March 2014, doi: 10.1109/JBHI.2013.2282125, PMID: 24058038
- Sujan Perera, Pablo Mendes, Amit Sheth, Krishnaprasad Thirunarayan, Adarsh Alex, Christopher Heid, Greg Mott, Implicit Entity Recognition in Clinical Documents, Proceedings of the Fourth Joint Conference on Lexical and Computational Semantics (*SEM) 2015, pp. 228-238.