Knowledge Based Fuzzy Inference System for diagnosis of Diffuse Goiter

A goiter, also spelled goitre, is a visible swelling of the neck or larynx resulting from enlargement of the thyroid gland and can be categorized into two main types: diffuse, nodular. Worldwide, over 90% cases of goiter are caused by iodine deficiency. Fuzzy inference systems (FIS) have, to date, proven to be extremely useful and efficient in the field of medicine and health care. To the best of our knowledge, no work has been carried out for diagnosis of goiter using FIS. In this paper, a novel, knowledge based FIS is developed for diagnosis of diffuse goiter using 2D ultrasound images employing Mamdani model for inferencing and centroid scheme for defuzzification. Features obtained using image processing are fuzzified. Further the inference rules are obtained using class based association rules. The outcome of proposed FIS accords with that of expert's opinion for almost all the tested cases.