GUI-BASED BRAIN TUMOR DETECTION

USING MACHINE LEARNING

Batch members
Under the Guidance of :
Mr. B. PRADEEP KHANTH M.E. ARAVINDHAN K - 513321106003

Teaching fellow ARTHI R - 513321106004

Department of ECE SUGI PRIYAN K - 513321106702

UCEA THANIGAIVEL M - 513321106304

Abstract:
 Our project aims to build an automated system for detecting brain tumors from MRI
images using Convolutional Neural Networks (CNN), improving speed and accuracy in
medical diagnosis.

A CNN model is trained to classify MRI scans into tumor and non-tumor categories by
learning deep features from the images.

A web-based, user-friendly interface is developed using Gradio, allowing users to upload

MRI images and receive instant predictions with ease.

 The system reduces manual workload, minimizes diagnostic errors, and makes
advanced AI tools accessible even to non-technical users in clinical environments.

 The integration of CNN with a Gradio-based GUI results in a fast, reliable, and practical
tool for early brain tumor detection.
Introduction:
Brain tumors are life-threatening conditions that require timely and
accurate diagnosis.

MRI (Magnetic Resonance Imaging) scans are commonly used for

detecting brain abnormalities.

Manual analysis of MRI images can be slow, complex, and susceptible to

human error.

Deep learning provides advanced solutions for image classification and

medical diagnosis.

Convolutional Neural Networks (CNNs) are particularly effective in

Introduction:
 CNNs can automatically learn features from MRI scans and classify them as
tumor or non-tumor.

 Our project uses a CNN model trained on brain MRI images to perform tumor
detection.

 To make the system more accessible, we created a Graphical User Interface

(GUI) using Gradio.

 The GUI allows users to upload MRI images and instantly view the model’s
prediction.

 This system aims to assist healthcare professionals by offering a fast, reliable, and
user-friendly tool for early brain tumor detection.
Objectives:
 To develop an automated system for accurate detection of brain tumors
from MRI images using deep learning.

 To design and train a Convolutional Neural Network (CNN) capable of

classifying MRI images into tumor and non-tumor categories.

 To simplify the user interaction by creating a clean and accessible

Graphical User Interface (GUI) using Gradio.

 To reduce the time and effort required for manual diagnosis and assist
medical professionals in faster decision-making.

 To improve diagnostic accuracy and minimize human error in the

interpretation of brain MRI scans.
Objectives:
To make the system usable even by non-technical users, such as
doctors and healthcare workers, through an intuitive interface.

To deploy the model efficiently so it can be used in real-time

scenarios with quick image upload and prediction.

To contribute to AI-powered healthcare tools that support early

diagnosis and better treatment planning.
Literature Survey:
S.N TITLE AUTHOR’S DESCRIPTION ADVANTAGE DISADVANTAG
O NAME S ES
1. Neurophysiological Jyotismita Reinforcement High Complexity of
Response Based Chaki, Marcin learning(RL) accuracy architecture.
on Auditory Sense Wozniak categorize the rates.
for Brain most
Modulation Using comparable
Monarual images to an
Beat(2023) unknown query
image.
2. Fluora-a System Elisabeth Klint, Fluroscene Improved Complexity and
for Combined Stina techniques tumor learning curve.
Fluroscene and Maturizon, identify tumor marginal
Microcirculation Johan Richter in the zone zone
Measurements in during detection.
Brain Tumor resection and
Surgery(2022) biopsy.
Literature Survey:
S.NO TITLE AUTHOR’S DESCRIPTION ADVANTAGES DISADVANTAG
NAME ES

3. A Robust Hasnain Ali Tumors Enhanced Dependence on

Approach for Brain Shah, Anand identified by efficiency, quality of MRI
Tumor Detection Paul, Faisal CNN models. So Image images,
in Magnetic Saeed it achieved the enhancement Overfitting.
Resonance images highest techniques.
Using Finetuned classification
EfficientNet accuracy.
(2022)
4. SEResU-Net for Chengdong SEResU-Net Enhanced Data imbalance,
Multimodal Brain yan, Hui can extract accuracy for Computational
Tumor Zhang, Jurong more feature small-scale cost.
Segmentation Ding information.It tumor.
(2022) segmenting
multimodal
brain tumors.
Literature Survey:
S.N TITLE AUTHOR’S DESCRIPTIO ADVANTAGE DISADVANTAG
O NAME N S ES

5. A Hybrid Feature Abdu Gumaei, The brain Improved Complexity of

Extraction Method Mohammed tumor classification feature
With Regularized Mehedi Hassan, features are accuracy, extraction,
Extreme Learning Abdulhameed extracted Effective Dependency on
Machine(RELM) for Aleliwi based on a preprocessing preprocessing
Brain Tumor hybrid . techniques.
Classification method of
(2019) feature
extraction.
RELM is used
for classifying
the type of
brain tumor.
Existing Work:
 Brain tumor poses a severe threat to millions, with a significant portion facing
low survival rates. Innovative solutions are crucial for early detection and
improved treatment outcomes.

 SVM-mRMRe (Support Vector Machine - minimum Redundancy Maximum

Relevance), offers a hybrid approach to feature selection in analysing
microarray data for Brain tumor and classification.

 XG Boost algorithm is used, which is an evolutionary lightweight model

intended for brain tumor detection.
 Makes use of magnetic resonance imaging (MRI) to analyse data, intensity,
texture, and shape are included in feature extraction.

 Useful for distinguishing brain tumor into four different grades provides a
Proposed Method:
 The Convolutional Neural Network(CNN) is to classify brain tumor
improving accessibility for medical practitioners.
 Application of magnetic resonance imaging (MRI) for brain tumor
detevtion.
 To ensure a smooth and engaging experience, GUI enables users to
upload medical images, start detection processes, and view results.
 In existing system, SVM and XGBOOST achieves accuracy percentage
ranging from 50% to 60%.
 When compared to existing system, the proposed system achieves
high accuracy ranging from 80% to 90% by CNN and KNN algorithms.
Proposed Method Design Flow:
CNN Algorithm:
 CNN is a deep learning model designed specifically for analyzing visual
data such as medical images.
 It uses convolutional layers to automatically extract relevant features
from input images.
 CNNs reduce the need for manual feature engineering by learning
patterns directly from the data.
 The architecture typically includes convolutional, pooling, and fully
connected layers for classification.
 CNNs are widely used in medical image analysis, including tumor
detection, due to their high accuracy and robustness.
Modified CNN Algorithm :
 We enhanced the traditional CNN by adjusting the number of layers
and filter sizes for better tumor feature extraction.
 Dropout and batch normalization techniques were added to improve
model generalization and prevent overfitting.
 The architecture was fine-tuned specifically for binary classification:
tumor vs. non-tumor.
 Early stopping and learning rate adjustments were applied to speed
up training and optimize performance.
 Our modified CNN achieved higher accuracy and better validation
performance compared to the standard model.
Modified CNN Algorithm Design
Flow:
How CNN Works
Convolutional[Conv2D]
Layer:
How CNN Works
ReLU

Max Pooling Layer:

Flattened Layer:
How CNN Works
Dense Layer

Dropout Layer
How CNN Works

Softmax Output
Working Methodology:
1. Data Collection & Preprocessing
• Brain MRI images are collected and preprocessed by resizing, normalizing, and
augmenting to improve model performance.

2. Model Design (Modified CNN)

• A customized Convolutional Neural Network is designed and trained to classify
MRI images into tumor and non-tumor categories.

3. Model Training & Evaluation

• The model is trained using labeled datasets, and its performance is evaluated
using accuracy, loss, and validation metrics.
Working Methodology:
4. GUI Development with Gradio
• A user-friendly interface is created using Gradio to allow users to upload
images and receive instant predictions.

5. Output & Interpretation

• The system analyzes the uploaded image, processes it through the model, and
displays the final result as “Tumor” or “No Tumor”.
Result :
Coding
Result :
Result :
Accuracy & Loss Plot Graph
User Dashboard (GUI):

Login page
The Login GUI page serves as the entry point to this site, ensuring secure
and controlled access to the system. It is designed with a clean and user-friendly
interface where users can enter their credentials to access the brain tumor
detection tool. Built using Python and Gradio, the login page enhances the
usability of the application while maintaining simplicity. It helps keep the system
Data Upload Page:
Result page:

Brain tumor Result (Normal)

Result page :

Brain tumor Result (Affected)

Key Features :
 User-Friendly GUI: Developed using Gradio for simple image upload and
instant results, even for non-technical users.
 Deep Learning Model (CNN): Utilizes a Convolutional Neural Network for
high-accuracy tumor classification.

 Real-Time Prediction: Processes and analyzes MRI images within seconds,

enabling faster diagnosis.

 Binary Classification Output: Clearly identifies whether a brain tumor is

present or not present.

 Visual Output with Confidence Score: Displays MRI image and

classification result along with prediction probability.

 Custom-Trained on Medical MRI Dataset: Ensures the model is fine-tuned

Application:
1. Medical Diagnosis Assistance:
• The GUI-based system serves as a valuable tool for healthcare professionals
in diagnosing brain tumor disease, providing accurate and efficient
classification using machine learning techniques.

2. Streamlined Workflow:
• By integrating a user-friendly GUI, the system simplifies the process of
uploading medical images, initiating classification, and visualizing results,
streamlining the diagnostic workflow for healthcare practitioners.

3. Enhanced Accessibility:
• The intuitive GUI interface caters to individuals with diverse technical
backgrounds, enabling healthcare professionals to easily access and utilize
advanced machine learning algorithms for disease classification without
Application:
4. Timely Intervention:
• The system facilitates early detection of brain tumor disease,
enabling timely intervention and treatment planning, which can
significantly improve patient outcomes and quality of life.

5. Improved Patient Care:

• By providing accurate and reliable classification results, the GUI-
based system empowers healthcare providers to make informed
decisions, leading to personalized treatment strategies and
enhanced patient care in neurological disorders.
Tools Used :

Hardware Software
PC(Personal Computer)  Google colab
 Python 3.10.6
 Gradio
Conclusion:
A GUI-based brain tumor classification system marks a major advancement
in medical imaging and diagnostics.
The intuitive interface enables healthcare professionals to easily access and
utilize machine learning tools.
MRI-based preprocessing and feature extraction enhance the system’s
ability to analyze complex neurological data.
The use of CNN algorithms significantly improves the accuracy and
reliability of tumor detection.
This system supports faster diagnosis, timely intervention, and better
clinical decision-making.
Ithas strong potential to revolutionize neurological diagnostics, leading to
improved healthcare and patient outcomes.
Future Scope :
In the future, the GUI-based system for brain tumor disease classification holds
considerable potential for expansion and refinement. It could evolve to incorporate multi-
modal imaging beyond MRI, enabling comprehensive diagnostic capabilities for various
neurological disorders. Additionally, advancements in feature extraction techniques tailored for
neuroimaging data, real-time diagnosis capabilities, and cloud-based deployment for remote
access and collaboration among healthcare professionals could further enhance its utility.
Longitudinal data analysis, integration of genomic and clinical data, and adaptation for
telemedicine applications represent promising avenues for future development. Continuous
optimization of machine learning algorithms and system parameters will be crucial to
improving performance, efficiency, and scalability over time, ensuring the system remains at
the forefront of neuroimaging analysis and diagnosis
References:
[1] M.A. Alsuwaiket, ‘‘Feature extraction of EEG signals for seizure detection using
machine learning algorthims,’’ Eng.Technol.Appl. Sci. Res., vol. 12, no.5, pp. 9247–
9251, Oct. 2022.

[2] P.Dhar,V.K. Garg, and M. A.Rahman, ‘‘Enhanced feature extractionbased CNN

approach for epileptic seizure detection from EEG signals,’’ J. Healthcare Eng., vol.
2022, Mar. 2022, Art. no. 3491828.

[3] S.Pattnaik,N.Rout, and S.Sabut, ‘‘Machine learning approach for epileptic seizure
detection using the tunable-Q wavelet transform based time– frequency features,’’
Int. J. Inf. Technol., vol. 14, no. 7, pp. 3495–3505, Dec. 2022.

[4] R.K.Joshi, V.M. Kumar, M.Agrawal, A.Rao, L.Mohan, M.Jayachandra, and

H.J.Pandya, ‘‘Spatiotemporal analysis of interictal EEG for automated seizure
References:
[5] C.O.Adetunji, O.T. Olaniyan, O.Adeyomoye, A.Dare, M.J.Adeniyi, and A. Enoch,
‘‘An intelligent diagnostic approach for epileptic seizure detection and classification
using machine learning,’’ in Artificial Intelligence for Neurological Disorders. New
York, NY, USA: Academic Press, 2023, pp. 225–243.

[6] M.S.Nafea and Z.H.Ismail, ‘‘Supervised machine learning and deep learning
techniques for epileptic seizure recognition using EEG signals— A systematic
literature review,’’ Bioengineering, vol. 9, no. 12, p. 781, 2022.
THANK YOU