Intelligent Decision Support System for Depression Diagnosis Based on Neuro-fuzzy-CBR Hybrid


  •  Victor E. Ekong    
  •  Udoinyang G. Inyang    
  •  Emmanuel A. Onibere    

Abstract

Depression disorder is common in primary care, but its diagnosis is complex and controversial due to the conflicting, overlapping and confusing nature of the multitude of symptoms, hence the need to retain cases in a case base and reuse effective previous solutions for current cases. This paper proposes a neuro-fuzzy-Case Base Reasoning (CBR) driven decision support system that utilizes solutions to previous cases in assisting physicians in the diagnosis of depression disorder. The system represents depression disorder with 25 symptoms grouped into five categories. Fuzzy logic provided a means for handling imprecise symptoms. Local similarity between the input cases and retrieved cases was achieved using the absolute deviation as the distance metric, while adaptive neuro-fuzzy inference system handled fuzzy rules whose antecedents are the mapped local similarities of each category of symptoms for global similarity measurement, upon which the retrieved cases are ranked. The 5 best matched cases are subjected to the emotional filter of the system for diagnostic decision making. This approach derives strengths from the hybridization since the tools are complementary to one another.



This work is licensed under a Creative Commons Attribution 4.0 License.