Authors :
Maurice Gatsinzi; Dr. Musoni Wilson
Volume/Issue :
Volume 9 - 2024, Issue 2 - February
Google Scholar :
http://tinyurl.com/5apytnj5
Scribd :
http://tinyurl.com/ytjerj7w
DOI :
https://doi.org/10.5281/zenodo.10691442
Abstract :
A brain tumor is a fatal condition that needs
to be surgically removed with precision. Brain cancers
were identified utilizing magnetic resonance imaging
(MRI). The goal of image segmentation for MRI brain
tumors is to create a distinct tumor boundary and to
identify the tumor area (also known as the region of
interest, or ROI) from the healthy brain. A brain tumor
is a malformed mass of tissue in which cells multiply
quickly and endlessly, unable to stop the tumor's growth.
Severe forms of cancer have a limited life expectancy,
often just a few months, in their most advanced stages.
MRI, CT, and ultrasound scans are a few types of image
modalities.
A brain tumor is a dangerous disorder brought on
by aberrant growth of brain cells. It is brought on by the
tissues encircling the skull or brain. Brain tumor
patients have aberrant mental states, vomiting,
headaches, seizures, trouble speaking and walking, and
eyesight impairments. The proliferation of malignant
cells Tissue is detected in magnetic resonance imaging
(MRI) images. We introduced a deep learning method
for identifying brain tumors in MRI data that is based
on the (Visual Geometry Group). VGG-16 Architecture.
The massive volumes of data produced by these
image scanning methods, however, make manual
analysis challenging and time-consuming. In this study,
deep learning techniques such as Convolutional Neural
Networks (CNNs) and Visual Geometry Group (VGG-
16) outperformed conventional methods in the
classification of brain cancers. The best deep features
from a variety of well-known CNNs and abstraction
levels were used in this fully automated computerized
approach to classify brain tumors. Transfer learning was
also applied to the images for the Visual Geometry
Group (VGG-16) change in relation to the number filters
in its feature training. In order to avoid over fitting, I
was analyzing the results after incorporating dropout
layers into this design.
Keywords :
MRI, CNN, VGG-16 and Brain Tumor.
A brain tumor is a fatal condition that needs
to be surgically removed with precision. Brain cancers
were identified utilizing magnetic resonance imaging
(MRI). The goal of image segmentation for MRI brain
tumors is to create a distinct tumor boundary and to
identify the tumor area (also known as the region of
interest, or ROI) from the healthy brain. A brain tumor
is a malformed mass of tissue in which cells multiply
quickly and endlessly, unable to stop the tumor's growth.
Severe forms of cancer have a limited life expectancy,
often just a few months, in their most advanced stages.
MRI, CT, and ultrasound scans are a few types of image
modalities.
A brain tumor is a dangerous disorder brought on
by aberrant growth of brain cells. It is brought on by the
tissues encircling the skull or brain. Brain tumor
patients have aberrant mental states, vomiting,
headaches, seizures, trouble speaking and walking, and
eyesight impairments. The proliferation of malignant
cells Tissue is detected in magnetic resonance imaging
(MRI) images. We introduced a deep learning method
for identifying brain tumors in MRI data that is based
on the (Visual Geometry Group). VGG-16 Architecture.
The massive volumes of data produced by these
image scanning methods, however, make manual
analysis challenging and time-consuming. In this study,
deep learning techniques such as Convolutional Neural
Networks (CNNs) and Visual Geometry Group (VGG-
16) outperformed conventional methods in the
classification of brain cancers. The best deep features
from a variety of well-known CNNs and abstraction
levels were used in this fully automated computerized
approach to classify brain tumors. Transfer learning was
also applied to the images for the Visual Geometry
Group (VGG-16) change in relation to the number filters
in its feature training. In order to avoid over fitting, I
was analyzing the results after incorporating dropout
layers into this design.
Keywords :
MRI, CNN, VGG-16 and Brain Tumor.
