| description abstract | This paper presents a novel decision support system (DSS) to assist medics administer optimal clinical diagnosis and effective healthcare post-treatment solutions. The DSS model that evolved from the research work predicted treatment of cerebral aneurysm using fuzzy classifications and neural network algorithms specific to patient clinical case data. The Lyapunov stability implemented with Levenberg–Marquardt model was used to advance DSS learning functional paradigms and algorithms in disease diagnosis to mimic specific patient disease conditions and symptoms. Thus, the patients' disease conditions were assigned fuzzy class dummy data to validate the DSS as a functional system in conformity with core sector standards of International Electrotechnical Commission—IEC61508. The disease conditions and symptoms inputted in the DSS simulated synaptic weights assigned linguistic variables defined as likely, unlikely, and very unlikely to represent clinical conditions to specific patient disease states. Furthermore, DSS simulation results correlated with clinical data to predict quantitative coil embolization packing densities required to limit aneurismal inflow, pressure residence time, and flow rate critical to design treatments required. The profiles of blood flow, hazards risks, safety thresholds, and coiling density requirements to reduce aneurismal inflow significantly at lower parent vessel flow rates was predicted by DSS and relates to specific anatomical and physiological parameters for post-treatment of cerebral aneurysm disease. | |
