TROPIX-MD |
||
| [Home][Main] | My Research Interests | |
| Diseases Diagnosed include:
Malaria Typhoid Fever Schistosomiasis Felariasis Helminthiasis Cholera Yellow Fever Hypatiasis Meningitis
|
TROPIX-MD - is
a medical expert systems software for diagnoses and treatment of some 22
typical tropical diseases.
TROPIX-MD was developed at the Center for Advanced Computer Studies, at The University of Southwestern Louisiana as part of the Ph.D. research work of Dr. A. Sunny Okorie. Several journal publications were produced from this work, including "Disease Diagnosis and Therapy using Case Based Reasoning", published in the International Journal of Artificial Intelligence in Medicine, 1997. Our goal at RNM is to use TROPIX-MD to assist the rural poor in rural and semi-urban clinics / Hospitals with no doctors, or limited health care workers. The management and screening of patients on a daily basis is a big chore and headache to the scanty doctors. The ratio of doctors to patient is 1-3000 in semi-rural areas and worse in more remote rural villages. The ability of a nurses aid, or health care worker to use TROPIX-MD to deal with the majority of cases presented daily is a big advantage. The diagnostic accuracy of TROPIX-MD from clinical tests and experimental usages was over 98%.
TROPIX-MD Research Abstract 1I have been studying some 22 tropical diseases with a view to developing an intelligent application tool that would support medical workers, specifically in rural and semi-urban communities of developing countries, especially Nigeria. There are very limited doctors and trained nurses. The turnaround time between when patients arrive in a hospital and when they see a doctor / treated is over 8 hours in most cases in most public hospitals / clinics. Medical resources, such as laboratory facilities are not available in many of the clinics / hospitals, or are ill equipped to diagnose problems.
Consequently, my research involves the use of the ideal-disease model representation in DM (Disease Matrix) of all known attributes of those diseases under study with some abductive reasoning in concert with case-based reasoning to validate diagnosis done with my Matched Vector Functions (MVF) approach. The MVF algorithm also incorporates the Singular Value Decomposition (SVD) technique - principal component analysis to find case weights which help generated some similarity ranking with more efficient case orthogonality. These similarity ranking of cases are then utilized on a Case-Base of past patient cases (Case-Based Reasoning) to validate the diagnosis reached with the MVF algorithm.
The TROPIX prototype has been built and demonstrated at CACS in Lafayette. However, final touches are being given to the application on the therapeutic side. In addition, clinical evaluation of TROPIX planned for late 1997 actually started in September of 1999. The interim results were impressive. However, a more rigorous evaluation in selected geographic zones is being arranged because it would provide a richer knowledge augmentation of the case base. The results when available would be used as input for further enhancement of the system prior to its deployment in the many more locations with limited health care workers or very high patient to doctor ratio (generally in excess of 5000 to 1). Hopefully, the system will be bundled together with Laptop PC's to rural hospitals and the targeted clinics as a decision support tool to relieve the pressures of medical workers, serve as a "blind" diagnostician to help the expert with second opinions, help clinics without doctors to run at a more reasonable pace than would otherwise be possible.
A. Sunny Ochi-Okorie was a Visiting Scholar, The CACS and Senior Lecturer (Associate Professor) at the FUTO in Nigeria. Currently, he is a Senior Analyst with the Shared Applications Group, Shell Information Technology, Houston - Texas; and serves as the President / General Manager of an International Relief & Development non-profit agency called Relief Network Ministries, Inc.
|
Dr. A. Sunny Okorie
1 This research was part of my Ph.D. work at the Federal University of Technology, Owerri (FUTO), with research bench work at the University of Louisiana at Lafayette’s Center for Advanced Computer Studies (CACS). It work was partly funded by the World Bank Staff Development program at the FUTO and the Nigerian Universities Commission, Abuja. Three journal papers / Conference proceedings, and two CACS technical reports have been published from this work. For more information, check the Tropix-md website at http://www.tropix-md.info
Research Interests
|
