An efficient approach for brain tumour detection based on modified region growing and neural network in MRI images

Brain tumor severity analysis using modified multi‐texton histogram and hybrid kernel SVM

“…When comparing the specificity the graph shows that the specificity of ak 1 , k 1 1 k 2 , k 1 *k 2 and which is greater than the existing works (Ahmed et al, 2011;Kavitha, 2012), and traditional method. On analyzing the accuracy the graph shows that the accuracy of our proposed work is 93% is better when compared with the existing works (Ahmed et al, 2011;Kavitha, 2012), and also with the traditional method. By analyzing the graphs, the specificity and sensitivity graphs show better results of our proposed modified MTH when compared with the existing work.…”

“…The average sensitivity of the proposed technique is 94.52% for MMTH1(k 1 1 k 2 ), 94.53% for MMTH1(k 1 *k 2 ), 94.50% for MMTH 1 ak 1 where the existing technique (Ahmed et al, 2011;Kavitha, 2012) achieved 25% and 28.5%, MMTH-based linear SVM achieved 37.8% sensitivity. The average specificity of the proposed technique is 74.52% for MMTH 1 (k 1 1 k 2 ), 74.53% for MMTH1 (k 1 *k 2 ), 74.50% for MMTH1 ak 1 where the existing technique (Ahmed et al, 2011;Kavitha, 2012) achieved 55% and 58% specificity. Conversely, MTH-based linear SVM has obtained only 45.5% sensitivity, but MTH-based linear SVM only 43.2%.…”

“…Here, with the aid of the input MRI image training and testing dataset, the values of TP, FP, FN, TN, sensitivity, specificity, and accuracy are given in Table II. Here, we have compared the proposed approach (modified MTH with hybrid kernels) with the traditional MTH (Ahmed et al 2011;Kavitha, 2012). From the Table II, our modified multi-texton histogram has a better performance compared with the multi-texton.…”

“…Comparative Analysis. We have compared our proposed modified multi-texton histogram based brain tumor detection method to existing techniques such as Kavitha (2012) and Ahmed et al (2011) and also with the traditional method. The performance analysis has been made by plotting the graphs of evaluation metrics such as sensitivity, specificity, and the accuracy.…”

Modified texture based region growing segmentation of MR brain images

“…For this, a threshold level for intensity value is set and those neighbour pixels that satisfy this threshold is selected for region growing. The normal region growing has two drawbacks: (1) noise or intensity variation leads to oversegmentation and (2) shading of the real image may not be distinguished [13].…”