Cardiac Image Analysis based on K Means Clustering
Download as PPTX, PDF1 like1,153 views
The document discusses segmentation techniques for cardiac image analysis, specifically K-means clustering. It describes the structure of the heart and need for cardiac image analysis to examine cardiac function and detect blockages. K-means clustering is introduced as a clustering method for segmentation that groups similar pixels into clusters by minimizing distances between cluster centers. The steps of the K-means clustering algorithm are outlined. Future improvements mentioned include automating the selection of the region of interest to improve stenosis detection.
Cardiac Image Analysis based on K Means Clustering
- 1. ASET Submitted By: NAVEEN TOKAS M.Tech ECE/2nd Sem ASET/AUH Guided By: Ms.Shruti K. ASET/AUH
- 2. ASET •Introduction •Segmentation Techniques •Introduction to K Means Clustering •Block Diagram •Future Scope
- 3. ASET •The heart is an essential organ of the human circulatory system. There should be a proper functioning of the heart to avoid Cardio Vascular Diseases (CVD). •In the cardiac image analysis we have to examine the cardiac function, structure in the left ventricle of the heart which is the main area for the stenosis or blockage. • Cardiac Image Analysis could be done with the help of one of the segmentation techniques known as K Means clustering. •The heart consists of four parts including left atrium, right atrium and the lower parts includes left and right ventricles. The most internal layer, i.e. the layer in contact with the blood, is called the endocardium the outer most layers, i.e. the layer which envelopes the heart, is called the epicardium. In between the endocardium and the epicardium is the myocardium, in which the muscular tissue responsible for the contraction of the heart. INTRODUCTION:
- 4. ASET Fig 01: Structure of Heart
- 5. ASET SEGMENTATION: •Segmentation is an operation that is used to portioning the images according to similarity, discontinuity or by determining the edges to analyze the information with the help of multiple partitions. It is an essential part that is used in medical image analysis. The different techniques under segmentation that used in the cardiac image analysis are categorized into following: •Thresholding Methods •Region Based Methods •Edge Based Methods •Clustering Methods
- 6. ASET • Clustering Method refers to the pixels having similar properties grouped together known as clusters. Grouping is based on maximizing the similarities, if inter class similarities is increase then the quantity of clusters are automatically increased to get the optimum results. •In the K Means Clustering the letter K is referred to the number of clusters that is to be decided in the starting of the algorithm. In this we have to define the K centers one for each cluster. The center should be far from the others so that the distance could be calculated easily and the data points could predict their clusters very precisely. Steps of K Means Clustering: •The number of clusters must be known to K. •The number of clusters center should be far from each other. •Consider each data point and assign it to the number of cluster which is closest. •Re-Calculate the cluster center by finding mean of the data points •Repeat third and forth step till shifting centers among the cluster could be observed.
- 7. ASET
- 8. ASET
- 9. ASET FUTURE SCOPE: •Every technique has its advantages or disadvantages like in K Means Clustering has an advantage of best implementation technique under segmentation but has an disadvantage of non automatically of chosen the ROI (Region of Interest) if this could be find automatically then the probability of stenosis would be less. •Cardiac Diseases always be a part of life and this method definitely helps the doctor to cure CV diseases through the MRI and CT scan means medical image analysis. It definitely help to calculate the parameters or analysis the data by the radiologist.