International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 06 | June 2024
p-ISSN: 2395-0072
www.irjet.net
BRAIN TUMOR DETECTION SYSTEM Gopika Subash Babu1, Prof. Dr. Sheeja Agustin2 1UG Student, Dept. of Computer Science and Engineering, APJ Abdul Kalam Technological University, Kerala, India 2Professor, Dept. of Computer Science and Engineering, APJ Abdul Kalam Technological University, Kerala, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Detecting brain tumors plays a vital role in
within clinical settings,assisting doctors in providing quick and precise diagnoses. By concentrating on the binary classification task—determining whether a tumor is present or not—this initiative strives to improve early detection, hence enhancing treatment outcomes and patient survival rates in neuro-oncology.
medical image analysis crucial for timely diagnosis and treatment planning. An inventive project is introduced here, merging Convolutional Neural Networks (CNNs) with machine learning to automate brain tumor detection. By utilizing Python libraries like TensorFlow, a comprehensive framework is built, integrating CNN-based feature extraction with conventional machine learning algorithms for binary tumor classification. The methodology includes pre-processing MRI scans, extracting significant features using a CNN structure, and inputting these features into machine learning classifiers to identify the presence or absence of a brain tumor. Extensive experimentation is conducted on a diverse dataset containing MRI images of brains with and without tumors to assess the performance of various CNN architectures and machine learning models. The results exhibit promising accuracy and efficiency in tumor detection tasks, with the developed framework achieving remarkable sensitivity and specificity rates. This system holds substantial potential in aiding healthcare professionals in precise diagnosis and treatment planning, ultimately enhancing patient outcomes in neuro-oncology.
1.1 Motivation
1. INTRODUCTION
Current techniques for identifying brain tumors heavily depend on the manual analysis of MRI scans by radiologists and healthcare experts. While this traditional method is somewhat effective, it comes with its own set of. These include the risk of human mistakes, subjective judgment, and the considerable time needed for a thorough evaluation. The manual assessment process may result in inconsistent findings, especially considering the intricate nature of brain tumors and their subtle display in imaging. Furthermore, the growing number of MRI scans in medical setups can overwhelm healthcare providers, causing delays in diagnosing and treating patients. Several current systems integrate basic image processing methods and computer-aided diagnosis (CAD) tools to support physicians. However, these systems often lack the advanced features necessary to precisely and effectively identify tumors.
Among the various organs present in the human body, the brain particularly stands out as being the most crucial. An issue commonly seen leading to dysfunction in the brain is the development of a brain tumor, characterized bythe uncontrolled growth of excess cells. These abnormal cells tend to consume essential nutrients that healthy brain cells and tissues require, ultimately resulting in brain failure. At present, doctors rely on manual examination of MRI scans to detect and evaluate brain tumors, a method known for its susceptibility to inaccuracies and inefficiencies. Brain cancer, being a severe and life-threatening condition, claims numerous lives due to delayed or incorrect diagnoses. The primary goal behind detecting brain tumors lies in enabling early diagnosis and treatment. This project aims to develop an automated system capable of identifying the presence of brain tumors in MRI scans. By utilizing CNNs, this system processes MRI images to determine the existence of a tumor, offering a reliable diagnostic tool for healthcare professionals. Such an automated system presents substantial advantages
One common method to detect brain tumors is Magnetic Resonance Imaging (MRI). However, the manual interpretation of MRI scans by radiologists can be time-consuming and subjective. This can lead to errors or delays in diagnosis. This project aims to develop an automated system using computer-based methods to identify brain tumors in MRI images. By utilizing Convolutional Neural Network (CNN) algorithms, the system can efficiently analyze MRI scans to determine the presence of a tumor. The process involves several stages: image preprocessing, feature extraction, and classification. During image preprocessing, MRI scans are enhanced to make it easier to identify relevant features. Featureextraction focuses on identifying significant characteristics within the images that indicate the presence of a tumor. In the classification stage, neural network techniques are utilized to determine if a tumor is present. This innovative approach aims to streamline the detection process and improve diagnostic accuracy in identifying brain tumors using MRI technology.
Key Words: Tumor, CNN, MRI, accuracy, detection
© 2024, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 879