Home About Login Current Archives Announcements Editorial Board
Submit Now For Authors Call for Submissions Statistics Contact
Home > Archives > Volume 20, No 8 (2022) > Article

DOI: 10.14704/nq.2022.20.8.NQ22268

Gliomas Analysis from Nervous System Using U-Net++ Segmentation with VGG-19 Net CNN Architecture

Para Rajesh, A.Punitha, P.ChandraSekhar Reddy


The majority of malignant brain tumours are gliomas, which are a form of CNS tumour. Gliomas are classified into four classes by the WHO depending on their aggressiveness. Low-grade gliomas (LGGs) are grade I-II gliomas, whereas high-grade gliomas (HGGs) are grade III-IV gliomas (HGGs). Before preceding tumor progression, proper categorization of HGGs and LGGs is critical for therapy selection. The foundation for glioma detection is MRI. However, owing to human interaction, examining MRI consume more time and it has more blunders. The paper proposes the segmentation based Gliomas analysis from nervous system using U-Net++ with VGG-19 Net CNN architecture. Here by segmenting the tumor by neural network based technique, the HGGs will be detected at the earliest. On the basis of pathology-proven 104 patients diagnosed with glioma, we trained and evaluated the models (fifty LGGs, fifty four HGGs). We examined the models using mean values of ninety eight percent accuracy, ninety six percent precision, 89.8 percent recall, and 86.2 percent F1- score. As per the findings of the experiments, our custom-designed U-Net++ VGG-19 Net_ CNN model performed as well as or better than the past models. The findings suggest that the proposed custom model is effective and reliable in categorisinggliomas into LGG and HGG.


Gliomas, CNS, LGGs, HGGs, U-Net++_VGG-19 Net_ CNN.

Full Text