An Efficient Region of Interest Detection and Segmentation in MRI Images Using Optimal ANFIS Network
| Autor: | S. P. Kaarmukilan, K Amal Thomas, Sucheta Biswas, Soumyajit Poddar |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Adaptive neuro fuzzy inference system
Computer science business.industry Feature extraction ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Probabilistic logic Pattern recognition ComputingMethodologies_PATTERNRECOGNITION Region of interest Classifier (linguistics) Median filter Segmentation Artificial intelligence Cluster analysis business |
| Zdroj: | Advances in Intelligent Systems and Computing ISBN: 9789811578335 |
| Popis: | The detection of tumour regions in Glioma brain images is a time-consuming task. This paper discusses the algorithm for efficient detection of tumour using Optimal Adaptive Neuro-Fuzzy Inference System (OANFIS). The proposed methodology consists of five modules: pre-processing, feature extraction, selection, classification and finally segmentation. The database are pre-processed with the help of median filter. Gray Level Co-occurrence Matrix (GLCM) features are extracted from both the images. To avoid the complexity, important features are selected using Crow Search Optimization (CSO) algorithm. In the next step, the shortlisted features are given to the ANFIS classifier to make classify an image as an abnormal or normal image. The ROI region of the abnormal images is separated with PFCM (Probabilistic Fuzzy C-means Clustering). The performance of the proposed methodology is analyzed in terms of sensitivity, accuracy and specificity. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|