Analysis of left atrium segmentation

and disease detection using

polynomial kernel

SUBMITTED BY
KAVITHA.G(720713104064) PROJECT GUIDE
KHOKHILA.P.A(720713104068) Mrs. MEHALA.G
NITHYAPRIYA.S(720713104099) ASSISTANT PROFESSOR
MOHANAPRIYA.S(720713104501)
 ABSTRACT
• We propose a joint information approach for automatic analysis of 2D echocardiography
(echo)data.

• The approach combines a priori images, their segmentations and patient diagnostic
information within a unified framework to determine various clinical parameters, such as
cardiac chamber volumes, and cardiac disease labels.

• The main idea behind the approach is to employ Analysis of both echo image intensity
information and corresponding segmentation labels to generate models that jointly describe
the image and label space of echo patients on multiple apical views.

• These models are then both used for segmentation and volume estimation of cardiac chambers
for detecting pathological abnormalities such as Mitral Regurgitation (MR).

• We report performance of the proposed approach in detection of Mitral Regurgitation severity

and also it classified patients that suffer from moderate or severe Mitral Regurgitation.
 Literature survey
• Left atrial size: physiologic determinants and clinical applications.
2006 Jun 20;47(12):2357-63.

• Echocardiogram view classification using low-level features

Published in: Biomedical Imaging (ISBI), 2013 IEEE 10th International Symposium
Date Added to IEEE Xplore: 15 July 2013

• A survey of shaped-based registration and segmentation techniques for cardiac images.

Article in Computer Vision and Image Understanding 117(9):966–989 · September 2013

• Echocardiography in Cardiovascular Public Health

Journal of the American Society of Echocardiography 22(6):649-56; quiz 751-2 .July 2009

• Assessment of Echocardiographic Left Atrial Size

Published By: JACC:Cardiovascular imaging Online ISSN:1876-7591 ,Published
online August 1, 2012.
 INTRODUCTION
• The blood flow through heart, transport oxygen, hormones, co2, blood cells and nutrients to
and from our body.

Mitral Regurgitation:
• Abnormal reversal of blood flow from left ventricle to left atrium.
• Leakage of blood backward through the mitral valve each time left ventricle contracts.
• May not have symptoms for decades.
 INTRODUCTION

Transthoracic echocardiography:
• The most common 2DE examination is Transthoracic Echocardiography(TTE).
• It is used to access the structure and functionality of heart.
• Techniques: B-mode
• Doppler flow imaging

a) Parasternal short-axis view (b) Parasternal long-axis view (a) Subcostal view (b) Subprastenal view
(PSAX). (PLAX).
 INTRODUCTION

• More commonly the apical and parasternal images are collected for analysis.
• Mainly to measure the volume of the left article and the mitral value to find the
mitral regurgitation.

(a) Apical 2- chamber view (b) Apical 3-chamber view

(A2C) (A3C)

(c) Apical 4-chamber view

(A4C)
 SYSTEM SPECIFICATION

HARDWARE REQIUREMNTS SOFTWARE REQIUREMNTS

The hardware used for the development of The software used for the development of the
the project is: project is:

PROCESSOR : Intel Pentium IV OPERATING SYSTEM : Windows 8

FRONT END : Matlab 7.12
RAM : 512 MB DOCUMENTATION :MS-Office
MONITOR : 15 inch color monitor
HARD DISK : 40 GB
FLOPPY DRIVE : 1.44 MB DATASET Link:
CD DRIVE : LG 52X www.osirix-viewer. com/resources/DICOM-
image-library/
KEYBOARD : Standard 102 Keys
MOUSE : 2 Buttons
 LIST OF MODULES

• Preprocessing phase
• LA segmentation
• Training
• Testing
• Volume estimation
• Disease SVM classifier
 DATA FLOW DIAGRAM

View
New Patient Training
preprocessing

LA Volume LA
Testing
Estimation segmentation

Disease
Detection
 MODULE I

Preprocessing phase:
• Preprocessing phase automatically determines the region of interest (ROI) defining
the anatomical location of the LA.
• We reduce the images from 800*600 pixels to 400*300 pixels to improve
computational efficiency and reduce memory requirements.
• For a new query image ROI is localized in this phase.
 MODULE I OUTPUT
Preprocessing phase:
 MODULE II
jica
• Independent component analysis (ICA) is a statistical and computational technique for
revealing hidden factors that underlying sets of random variables, measurements, or signals.

• ICA defines a generative model for the observed multivariate data, which is typically given
as a large database of samples.

• These independent components, also called sources or factors, can be found by ICA.

• ICA is a much more powerful technique, however, capable of finding the underlying factors
or sources when these classic methods fail completely.
MODULE II OUTPUT
 CONCLUSION

• This project propose a complete program joint multi-see division of LA

and MR examination using B-mode intensity data.

• Advantages of this joint approach

 Evaluate the LA volume .
 Make understand particular models equipped for recognizing MR
names.