IOSR Journal of Nursing and Health Science (IOSR-JNHS) e-ISSN: 2320–1959.p- ISSN: 2320–1940 Volume 7, Issue 6 Ver. VIII. (Nov.-Dec.2018), PP 42-52 www.iosrjournals.org DOI: 10.9790/1959-0706084252 www.iosrjournals.org 42 | Page Differential Diagnosis of Eye Diseases Based on Fuzzy Cognitive Map 1 Okure Udo Obot, Iberedem Ibibio Udo and Samuel Sunday Udoh Department of Computer Science,University of Uyo, Uyo. Nigeria Corresponding author;Okure Udo Obot Abstract: The prevalence of eye diseases and its attendant effects which lead to blindness motivated an investigation into how diagnoses are undertaken. One startling revelation is the dearth of optical experts to cope with the increasing cases of eye diseases in clinics. This further prompted the design of a differential diagnosis tool for the purpose of aiding opticians in their assignment of diagnosis of eye diseases. Conflicting symptoms cause a lot of confusion to medical experts especially the inexperienced ones. In an attempt to carry out this task, two opticians were consulted independent of one another to assist in weighting 17 symptoms of 6 confusable eye diseases. The symptoms were weighted with respect to their causal relationship to one another and the 6 diseases. In addition to this, the opticians assisted in giving diagnostic results of 20 hypothetical cases independent of one another. Armed with this information and datasets, a Fuzzy Cognitive Map (FCM) was designed for the 17 symptoms and 6 diseases. From the FCM, an adjacency matrix was constructed. To obtain optimal weights through the links of the maps, a training session was undertaken using Hebbian learning rule. With all these, the 20 hypothetical cases were subjected to train the map. Results obtained from the exercise were compared with that obtained from the two opticians. A correlation of 0.65 was observed with the results of the first optician while a correlation of 0.45 was seen from that of the second optician. A novel feature of the differential diagnosis using the FCM methodology is the ability to give varying degrees of diagnosis of the main ailment and in addition that of an associated ailment. In doing so, the optician who is the ultimate user of the system could choose which of them to treat first maybe considering the severity of the ailment and how life threatening is one to another. Keywords: Fuzzy logic, Cognitive map, Differential diagnosis, Hebbian Learning, Eye Diseases. --------------------------------------------------------------------------------------------------------------------------------------- Date of Submission: 29-11-2018 Date of acceptance: 12-12-2018 --------------------------------------------------------------------------------------------------------------------------------------- I. Introduction The eye is one of the vital organs of the body and serves as an organ of sight. Its position in the body exposes it to so many dangers making it vulnerable to diseases. A diseased eye is a nightmare to the body and one of the major causes of regular visits to the hospitals. Ironically, most times the visit is made at an advanced stage of the disease when self medications had failed. Misdiagnosis of the eye diseases due to conflicting symptoms often lead to complication of the problem and often to blindness. In some cases it causes permanent problem that requires the use of eye-glasses all through the life time of the patient. The inability of an optician to differentiate one symptom of eye diseases from the other is further complicated by the inability of the patient to explain to the optician exactly how he (the patient) feels. The ambiguity in expression leads the optician to confusion and subsequently misdiagnosis of the disease. Some misdiagnoses complicate the problems of the patient as such diagnoses will cause wrong therapy for the patient. The therapy rather than relief the pains could cause grief. Medical diagnostic system deals with eliciting both vital and non-vital information from the patient which the medical doctor will try to decipher such information and relate with the structured knowledge of the suspected disease. Aside from this information, the laboratory results can also be used to establish a relationship of the symptoms and the suspected disease. In effect, there are relationship and interrelationship with causal influences. All these assist the medical doctor in taking a decision about a patient. The cognitive ability of the medical doctor to map the signs and symptoms elicited from the patients as the causal effect of the suspected disease enable the medical doctor to confirm the presence or absence of the disease in the patient. Cognitive maps according to Eden (2004) are collection of nodes linked by some arcs or edges. The nodes represent concepts or variables relevant to a given domain. A causal relationship is established between the nodes through the arcs or edges. The cause could be positive (excitatory) or negative (inhibitory) and where there is no causal effect the link represents null or zero. In the positive effect, the source node is said to excite the target node meaning that an increase in the value of the source node leads to a corresponding increase in the value of the target node. Corollary, in the negative effect the source is said to inhibit the target node, showing