Brain Tumor Detection with Hybrid Segmentation using Improved Adaptive Network Fuzzy Inference System

In the past, numerous researchers used a variety of cutting-edge deep learning models to work on biomedical image classification and segmentation systems. The results of deep learning are better than those of the current conventional computer vision methods. Currently various automated image segmentation methods have been used for feature extraction. These methods generally use a brain tumor classification approach and color detection schemes to predict the class label of neighboring pixels and the classification by validating the extracted features. These techniques' parameter improvement is dependent upon the signal processing approaches. The accuracy has increased significantly as a result of the usage of optimal parameter sets by numerous researchers, which are produced for a particular dataset. On the other hand, the model does not scale on any other collection of data. We have presented a novel hybrid approach for brain tumor segmentation that combines wavelet and watershed transform. This hybrid approach can be applied to a wide range of image types. When the model is used on a variety of images, the accuracy improves even when noise is present.