Опубликована: Янв. 1, 2024
Язык: Английский
Опубликована: Янв. 1, 2024
Язык: Английский
Egyptian Informatics Journal, Год журнала: 2024, Номер 29, С. 100596 - 100596
Опубликована: Дек. 20, 2024
Язык: Английский
Процитировано
0Опубликована: Июнь 5, 2024
Brain tumor being one of the major health hazards always needs a prompt diagnosis for early treatment options to improve chances survival patients. Traditional manual assessment MRI (magnetic resonance imaging) identify these conditions is common practice. Hence, automation practices can considerably quality procedures. In recent years almost every method automated brain detection uses deep learning technique. Despite decent works current techniques need significant improvement produce better results. Along with failure results, models present very complex architectures which sometimes require huge computational resources. Therefore, in this paper we novel that lightweight dual-stream model dual-input detect tumors on images. Both inputs use different pre-processing mechanisms based Contrast Limited Adaptive Histogram Equalization (CLAHE) and White Patch Retinex algorithm order enhance feature capabilities proposed model. addition, two-fold margin loss boost training process learning. The produces state-of-the-art results Glioma, Meningioma Pituitary tumors. presents accuracy 99%, 98% 99% respectively detection.
Язык: Английский
Процитировано
0Опубликована: Янв. 1, 2024
Язык: Английский
Процитировано
0