AUTHOR=Ali Tahir Mohammad , Nawaz Ali , Ur Rehman Attique , Ahmad Rana Zeeshan , Javed Abdul Rehman , Gadekallu Thippa Reddy , Chen Chin-Ling , Wu Chih-Ming 

TITLE=A Sequential Machine Learning-cum-Attention Mechanism for Effective Segmentation of Brain Tumor

JOURNAL=Frontiers in Oncology

VOLUME=Volume 12 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2022.873268

DOI=10.3389/fonc.2022.873268

ISSN=2234-943X

ABSTRACT=A brain tumor is the spread of abnormal tissues in the brain. There are about 120 types of brain tumors. Glioma is the most common type of tumor that is difficult to detect with the lowest survival rate. Magnetic Resonance Imaging is the most generally utilized imaging methodology that permits radiologists to look inside the cerebrum using radio waves and magnets; however, the manual identification of the tumor region is a tedious task. Therefore, reliable and automatic segmentation and prediction are necessary for the segmentation of brain tumors. Meanwhile, it is tedious and complex to identify the tumorous and non-tumorous regions due to the complexity in the tumorous region. This paper proposes a reliable and efficient neural network variant, i.e., an attention-based convolutional neural network for brain tumor segmentation. Specifically, an encoder part of the UNET is a pretrained VGG19 network followed by the adjacent decoder parts with an attention gate for segmentation noise induction and denoising mechanism for avoiding overfitting. The dataset we are using for segmentation is BRATS'20 which comprises four different MRI modalities and one target mask file. Above mentioned algorithm resulted in a dice coefficient of 0.83, 0.86, and 0.90 for enhancing core and whole tumor, respectively.