Classification and Identification of the constituent Elements of Human body Tissues using the Nearest Neighborhoodpoint model and Adaptive Neural Fuzzy Inference System

This paper introduces a novel artificial intelligence approach for identifying tissueelements within the human body using X-ray imaging which could be used in medicalfields such as diagnosing cancer, tissue damage, food, and also in nuclear agricultureto eliminate pests and detect the amount of toxins remaining in plants, etc. Theproposed method combines two artifitial intelligent techniques: the Adaptive NeuralFuzzy Inference System (ANFIS) and K-Nearest Neighbor (KNN) classification,implemented in MATLAB, to analyze data generated from Monte Carlo simulationsusing the GEANT۴ code. ANFIS, capable of adjusting fuzzy rules based on datavariations and environmental factors, is complemented by KNN classification, whichpartitions data into distinct groups based on point distances. This integrated approachenables the calculation of photon flux within tissues of varying sizes, facilitating accurateidentification of tissue constituents. The efficacy of the proposed method wasevaluated through simulations conducted on various tissues, including Skin, Kidney,Liver, Breast, Brain, Eye Lens, and Lung, within the GEANT۴ environment. Resultsdemonstrate superior accuracy and efficiency of the proposed method, with minimalprediction errors.

کلیدواژه ها:

ANFIS ، Element ، Detection ، GEANT۴ ، KNN