Analysis of the Performance of Classifiers on Wavelet Features with PCA and GA for the Detection of Breast Cancer in Ultrasound Images
Journal Title: IOSR journal of VLSI and Signal Processing - Year 2018, Vol 8, Issue 1
Abstract
Breast cancer is the most commonly diagnosed life threatening cancer in women worldwide. Breast cancer is the leading cause of cancer death among women. Early detection is of great significance and essential to the treatment of breast cancer. Ultrasonography is one of the most widespread imaging modality used to detect and classify abnormalities of the breast. This paper proposes the use of wavelet transform and its coefficients as texture features for the detection of abnormalities in the breast. Gray level co-occurrence matrix is computed from wavelet coefficients at two levels. Principal component analysis and genetic algorithms are used for feature reduction and selection. Support vector machine (SVM) and Naïve Bayes (NB) are used to differentiate benign and malignant lesions. Their performances are evaluated using diagnostic accuracy, sensitivity, specificity, positive predictive value, negative predictive value and Mathew’s correlation coefficient. The proposed method results in high classification accuracy of 98.57% in a data set containing 70 (30 benign and 40 malignant) breast ultrasound images. Results indicate that the proposed features can effectively characterize the properties of breast lesions in ultrasound images.
Authors and Affiliations
Nanda S1 ,, Sukumar M2
Keywords
Related Articles
- EP ID EP412551
- DOI -
- Views 161
- Downloads 0
A Review on Recognition of Online Handwriting in Different Scripts
Online handwriting recognition or character recognition is the process in which a handwritten message is recognized by processing the handwritten data. It is the process of converting handwritten characters to machine fo...
Study of High Performance of Gan-Based HEMT Having Two Channel Layers of Gan/Inalgan
In this paper, the impact of GaN-based High Electron Mobility Transistor (HEMT) is reported. The device is having two channel layers of GaN (Gallium Nitride) and InAlGaN (Indium Aluminium Gallium Nitride). The device is...
Support Vector Machine Classification of Stress Types in Speech
Speech of human beings is the reflection of the state of mind. Proper evaluation of these speech signals into stress types is necessary in order to ensure that the person is in a healthy state of mind. In this work we pr...
A Study onFace Detection and Image Steganographyusing Discrete Wavelet Transform
There are many methods available for secret data hiding and transfer, among which steganography stands out in various aspects. Steganography is used for secret data transfer in a cover object which may be an image, video...
Performance Analysis of FINFET Based 16:4 Encoder and 4:16 Decoder-A Design Perspective
The growing technology for the silicon chip is becoming a demand day by day to ensure the optimization in area, power consumption and speed in the digital circuits. The CMOS technology is fast growing domain where the an...