Causality: models, reasoning, and inference
Causality: models, reasoning, and inference
Inside Case-Based Reasoning
| Hi-index | 0.00 |
This paper describes work in progress investigating the feasibility of enhancing the treatment of diabetes through a web-based case-similarity retrieval system that permits physicians to pool their knowledge and optimize treatment of their patients. Diabetes mellitus type 2 is simple to diagnose but treatment is fraught with complications. Doctors often rely on their experiences with "similar" patients --- typically those having a combination of common factors. If doctors had a greater experience pool to draw from and could access the data in a manner that simulates their professional expertise, better treatment recommendations may result. The project involves developing a prototype consisting of a back-end database, a similarity and reasoning component, visualization techniques, professional interaction tools, and an end-user web interface. The system provides functionality for collecting and managing patient information while protecting patient privacy and for determining which cases in the database should be considered similar to a new case. The architecture is designed to permit future adaptation capabilities, such as considering the results most often selected or favored by doctors and self-adjusting after having determined which factors doctors actually pay most attention to. This paper includes an overview of the domain and specific problems addressed, a description of the research methodology and initial architecture, and initial outcomes that resulted from prototype development. It provides insights that may prove useful for others who are contemplating engaging in this type of project or who may be interested in contributing/benefiting from this specific effort. This research was conducted by undergraduates with a faculty mentor as part of the Collaborative Research Environment for Undergraduates in Computer Science and Engineering program sponsored by the Computing Research Associaion Committee on the Status of Women in Computing Research (CRA-W).