مؤسسة الشرق الأوسط للنشر العلمي
عادةً ما يتم الرد في غضون خمس دقائق
Brain lesion detection has an important role for human health care. People with Alzheimer's or traumatic brain injuries have lesions in the brain area. In this research, brain lesions were detected by the use of extreme learning based on deep learning. To improve the Extreme Learning Machine and increase accuracy. The Fully Connected Layer was replaced with an Extreme Learning Machine. The weights of Convolutional Layer and First Layer of extreme learning Machine Optimized with Imperialist Competitive algorithm. The ISLES 2015 dataset is used to calculate the accuracy of the proposed method. The results show that the value of DSC (Dice Similarity Coefficient) is near 91.3%