A fuzzy interval model for assessing patient status and treatment effectiveness using blood morphology

This study explores the generalization of heterogeneous medical data for monitoring anomalies and changes over time using fuzzy intervals. The most important feature of these intervals is saving the parameter value as a membership function from the interval [0, 1]. An example illustrating this method is the blood count parameters of an oncological patient recorded for three years with a monthly frequency. Over 20 typical measurements of these features are considered, and eight with the highest variance are selected. The registration of the overall picture of changes, a synthesis of eight fuzzy intervals, allowed for observing a systematic improvement in health. This approach allows the doctor to take a holistic view of the patient’s health (based on blood tests), avoiding the dilemma of which parameters are less and which are more important. The Mamdani fuzzy inference system was used to assess the patient’s health status. The study presents the actual results of medical measurements, and the GitHub repository contains measurement data