You are on page 1of 24

SRIDEVI WOMEN’S ENGINEERING COLLEGE

Department of Computer Science And Engineering


Accredited NBA & NAAC
Affiliated to JNTUH, Approved by AICTE, New Delhi
V.N.Pally, Near Gandipet, R.R.Dist-5000075
2022-2023

Identifying Bone Tumor Using X-Ray Images With Deep Learning .

Coordinators : Under the Guidance of:


Dr.V.Anantha Krishna Mrs.S.Mounasri(Asst.Prof)
Mr.B.Suman(Asst.Prof)

Presented By:Batch A13


A.Mounika-19D21A0505
Gurpreet Kour - 19D21A0522
P.Kavya Reddy - 19D21A0545
CONTENTS:
 Abstract
 Introduction
 Existing method
 Disadvantages
 Problem statement
 Proposed system
 Advantages
 System Specifications
 Functional And Non Functional Requirements
 System Design Architecture
 Algorithm
 Modules
 Data Flow Diagrams or UML Diagrams
 Conclusion
ABSTRACT:
The main objective of this project is to analyze and determine bone tumor disease of a
patient with X ray images using Convolutional Neural Network(CNN).A quick and
preliminary diagnosis of bone cancer begins with regular X ray. In this project an
ensemble technique has been proposed for successful delineation of tumors into intra-
tumoral classes. This technique improves the accuracy of identifying bone tumor on the
task of diagnosis.
INTRODUCTION:
A bone tumor is an abnormal growth of tissue in bone which leads to bone cancer.
Based on certain radiographic features, benign and malignant bone tumors have a
wide variety of appearances. Because a significant number of bone lesions are
uncommon or rare, only a limited number of radiologists develop sufficient
expertise to diagnose bone lesions and tumors accurately. We extract certain
features from bone X-ray images and use support vector machine (SVM) to
discriminate healthy and cancerous bones.
EXISTING METHOD:
 The existing system for diagnosis of bone tumor is regular X ray image which
makes use ionizing radiation.
 It is the process of imaging separating abnormal tissues from normal tissue.
 X-ray image analysis provides one of the primary screening tools for the
diagnosis of bone cancer.
 Here we use support vector machine(SVM), SVM algorithm finds the closest
point of the lines from both the classes. These points are called support vectors.
 “Support Vector Machine” (SVM) is a supervised machine learning algorithm that
can be used for both classification or regression challenges.
 The linear SVM model had a classification accuracy of 90.6% with 10 transcript
variables and no specificity is maintained.
DISADVANTAGES:
 Prediction and accuracy is less.
 A small increase in the possibility that a person exposed to X-rays may develop
cancer later in life.
 Consumes more time.
PROBLEM STATEMENT:
 As the identifying of bone tumor is not much accurate by using existing
system. How these accuracy of the description of the bone tumor is obtained is
seen in the proposed system. This project describes about the advanced
technique of Identifying Bone Tumor Using X-Ray Images with the help of deep
learning which is previously set to be regular X ray.
PROPOSED SYSTEM:
 In this project, a method is introduced to detect bone cancer by using deep learning
algorithm.
 The proposed system uses algorithm such as CNN which is a type of Artificial Neural
Network which comes under Deep learning.
 A CNN uses a system which is like a multilayer view-point that has been designed for
reduced process necessities.
 This model may achieve 90% accuracy and 88% specificity in the classification of bone
tumors as malignant or benign.
ADVANTAGES:
 The resulting method is very fast.
 It improves accuracy of prediction compared to existing system.
 Robust.
SYSTEM REQUIREMENTS SPECIFICATIONS
1.HARDWARE REQUIREMENTS:
 Processor - intel i3 or above
 Speed - 1.1 Ghz
 RAM - 4 GB or above
 Hard Disk - 500 GB or above
 SOFTWARE REQUIREMENTS:
 Operating System - Windows10 or above
 Programming Language - Python
FUNCTIONAL REQUIREMENTS:
Data collection.
Data segmentation.
Classification.
Testing.
NON FUNCTIONAL REQUIREMENTS:
 Performance.
 Accuracy.
 Maintainability.
 Traceability.
 Reliability.
SYSTEM DESIGN ARCHITECTURE:
ALGORITHM:
 Here we use CNN algorithm which is a type of Artificial Neural Network which
comes under Deep learning.
 A CNN can have multiple layers, each of which learns to detect the different
features of an input image. A filter or kernel is applied to each image to produce
an output that gets progressively better and more detailed after each layer. In the
lower layers, the filters can start as simple features.
MODULES:
DATASET UPLOAD
DATASET PRE-PROCESSING
FEATURES EXTRACTION
SEGMENTATION AND CLASSIFICATION
MODULES:
To implement this project we have designed following modules
1. Dataset upload: using this module we will upload dataset to application
2. Dataset Pre-processing : using this module we will read all images and then convert
them into GREY format and then resize all images to equal size and then normalize
pixel values.
3. Features Extraction: features or pixel values will be extracted from processed images
and then input this features to CNN to trained tumor prediction model
4. Segmentation & Classification: using this module we will read test image and then
apply segmentation to extract tumor part and then predict weather image is normal
or contains any tumor. Edge detection technique will be applied to surround
bounding box across tumor part
UML DIAGRAMS:
CLASS DIAGRAM:
USE CASE DIAGRAM:
SEQUENCE DIAGRAM:
DEPLOYMENT DIAGRAM:
COMPONENT DIAGRAM:
ACTIVITY DIAGRAM:
CONCLUSION:
• This project offers an automated analysis and prediction of bone tumor using
CNN algorithm.
• This ensemble method successfully generates highly accurate prediction of bone
tumor compared to existing methods.
• Most of the times in detection of tumor the images obtained by regular X ray
doesn’t give whether the tumor is malignant or not.
• Hence, an approach have been proposed which will comprised of number of
stages ultimately leads to the proper detection of tumor.
THANK YOU

You might also like