| Abstrakt: |
Accurate diagnosis of diseases and their level of intensity is a key aspect of health. A minor error in medical diagnosis, besides diverting the course of treatment and increasing the duration of therapy, raises the cost of services, leads to psychological trauma, and most importantly jeopardizes patient safety. The shorter the time of diagnosis, the easier it is to prevent, fight, treat and control the disease at lower intensity levels. Thus, it is essential to provide a model to deal with the possible errors of diagnosis in healthcare system. The present article introduces an engineering method for detecting and diagnosing faults in patient treatment. In the first phase, the factors affecting the emergence of the disease are predicted using Perceptron neural network. In the second stage residual values are produced; and in the third stage, in case there is a fault, the sources and causes of the fault are examined with the help of statistical quality control charts. Validation of the proposed model is evaluated by providing a numerical example. [ABSTRACT FROM AUTHOR] |
