Iranian Journal of Biomedical Engineering (IJBME)
نوع مقاله : مقاله کامل پژوهشی
نویسندگان
1 گروه مهندسی کامپیوتر، واحد قوچان، دانشگاه آزاد اسلامی، قوچان، ایران
2 گروه پرستاری، واحد قوچان، دانشگاه آزاد اسلامی، قوچان، ایران
کلیدواژهها
عنوان مقاله English
نویسندگان English
Accurate and rapid brain tumors diagnosis of brain tumors in magnetic resonance imaging (MRI) is one of the major challenges in the health and medical fields, especially in radiology. The deep learning models have been widely used in the medical field due to their high capabilities in extracting complex and nonlinear features from image data. However, deep learning models may have unstable results due to their strong dependence on model settings and sensitivity to input data quality. In this study, in order to improve the accuracy and precision in diagnosis of brain cancer tumors, a deep learning model is proposed based on two swarm intelligence algorithms, in which the particle swarm algorithm is employed to adjust the hyper parameters of the deep neural network and the ant colony algorithm is applied to select the most effective features of MRI images. Experimental results of tests on the Kaggle MRI brain tumor as benchmark dataset indicate 95% accuracy and 91% accuracy of brain tumor diagnosis obtained from the proposed model in this study, as well as improved performance compared to conventional deep learning methods.
کلیدواژهها English
