Accurate decision support systems are central to the medical prevention, diagnosis, and treatment of diseases such as autism, cancer, celiac disease, and lobar pneumonia. There are medical decision support systems that apply the properties of fuzzy logic and neural networks to assist physicians to minimize medical errors have had limited success. But how should effectual decision support systems be designed and realized to aid physicians in curtailing medical errors? Amirkhani et al. present a nomenclature, techniques, and applications of fuzzy cognitive maps that are used in present-day medical decision support systems.

A fuzzy cognitive map (FCM) is an uncertain directed graph for exhibiting the computational requirements and intricacies of a system. FCMs use a mixture of the topographies of neural networks and fuzzy logic to illuminate multifaceted systems. Specifically, to construct an FCM, a user (1) chooses the total and category of model notions, (2) defines the first model weight and the associations and connections among concepts, and (3) uses learning procedures to train the early weight to attain the ultimate model. The authors clearly show how an FCM model is used to classify celiac disease.

The medical applications of FCMs are characterized into managerial, diagnosis, forecast, and medicinal classification areas. Clinicians, as managers, can use FCM models to explore the natures of diseases and recommend treatments for patients. Doctors can use the FCM models to accurately distinguish diseases with similar symptoms. Knowledgeable physicians should be able to use an FCM system to improve therapeutic measures for curing progressive diseases.

Are you still searching for a comprehensive history of the models, embedded software, and alternative weight learning algorithms of FCMs? If so, I encourage you to read the insightful ideas in this paper for creating effective medical decision support systems that mimic human minds. The FCM models and algorithms support reliable sensitivity analysis for discovering alternative solutions to a variety of many medical what-if questions. I strongly encourage all artificial intelligence experts to read this paper and participate in the ongoing intellectual contributions of FCMs to medicine.