Classification of Kidney Images Using Cuckoo Search Algorithm and Artificial Neural Network
ABSTARCT :
Ultrasound (US) imaging is used to provide the structural abnormalities like stones, infections and cysts for kidney diagnosis and also produces information about kidney functions.
The goal of this work is to classify the kidney images using US according to relevant features selection. In this work, images of a kidney are classified as abnormal images by preprocessing (i.e. grey-scale conversion), generate region-ofinterest, extracting the features as multi-scale wavelet-based Gabor method, Cuckoo Search (CS) for optimization and Artificial Neural Network (ANN).
The CS-ANN method is simulated on the platform of MATLAB and these results are evaluated and contrasted. The outcome of these results proved that the CS-ANNN had 100% specificity and 94% accuracy. By comparing it with the existing methods, the CS-ANN achieved 0% false-acceptance rate.
Keywords: Kidney diagnosis, Gabor feature extraction, Cuckoo search, Artificial Neural Network, Ultrasound images.
EXISTING SYSTEM :
An automated system is developed for the diagnosis of kidney diseases by using ultrasonic systems in the recent years. During imaging, the system allows the extraction of vast data and good quality of information to detect the diseases.
The evaluation of global conditions can be made by the process of feature extraction, analysis of images and classify the images by pattern recognition techniques. But still, no technique had improved the accuracy of the system or proved to be best in accuracy for classifying the kidney diseases.
Therefore, the classification accuracy will be improved with the help of improving the existing pre-processing as well as classification models. This above mentioned limitations and the lack of solutions motivated this research work
DISADVANTAGE :
The scanning of US is used to scan the organs of the body like gallbladder, bladder, liver, ovaries, spleen, kidneys, pancreas, uterus and fetus in pregnant patients [4,5]. The complex geometric problems are solved by some effective algorithms like three-dimensional (3D) modelling, sharp extraction and classification method in medical image processing.
PROPOSED SYSTEM :
Deep learning is a subset of machine learning and artificial intelligence in which algorithms are developed to train machines/computers to detect a specific feature or classify objects. There is limited assessment of this method to identify and classify images in urology. We sought to assess the accuracy of deep learning method to automatically detect stone composition from images of kidney stones.
Deep learning computer vision methods can be used to detect kidney stone composition with high accuracy and has the potential to replace laboratory analysis of stone composition. The accuracy can be improved by increasing the number of stones used for training the network. Future work is needed to see if deep learning can be used for detecting mixed kidney stone composition.
ADVANTAGE :
The diagnosis of a disease can be affected by any error in medical imaging. Hence, the difficult task in medical field is to find the accurate identification of images for classification. The medical examinations and numerous data are collected from clinical trials for ensuring the statistical significance of studies.
|