A modified select and test (ST) algorithm for medical diagnosis in an ad-hoc network environment

Much of medical diagnosis is now being either assisted or fully carried out by expert systems. These expert systems are driven by reasoning algorithms such as backward and forward reasoners, select and test (ST), and many others. ST proofs to be more accurate compared to other medical reasoning algorithms, and this is achieved by its adoption of different reasoning patterns combined into the ST algorithm. However, ST model derived from the algorithm reasoning structures are not very optimized for better efficiency of medical domain knowledge and diagnostic accuracy. Hence, this paper proposes to improve these two deficiencies: efficiency and accuracy of ST model by incorporating the concept of the semantic web through the use of reasoners, coordinated rule set, and ontological-based knowledge base. Also, in this paper, all the contributory features or sub-modules of the improved ST algorithm are presented and their weight discussed. Furthermore, the improved ST algorithm is deployed in an ad-hoc network environment to pave way for medical experts to provide checking mechanism for consistency and coherency of result of the improved ST algorithm with that medically acceptable based on clinical protocols. The resulting ST algorithm is currently being fully implemented as a dedicated medical expert system for diagnosing breast cancer as a test bed.