Detection Of Abnormal Liver In Ultrasonic Images From Fcm Features

Document Type : Primary Research paper


1 Professor, Department of Electronics and Communication Engineering, Bannari Amman Institute of Technology, Sathyamangalam, India

2 Assistant Professor, Electronics and Instrumentation Engineering, Bannari Amman Institute of Technology, Sathyamangalam, India


The objective of this paper is to detect the liver abnormalities from ultrasonic
image database. The liver cancer is one of the major health issuesnow a day. Medical
Imaging techniques are proposed to diagnose the abnormalities in the earlier stage. In this
paper extracted Fuzzy C means (FCM) Clustering features of liver images and five
classifiers like Expectation maximization, Gaussian Mixture Model, Linear Discriminant
Analysis,Bayesian Linear Discriminant Analysis Classifier, Logistic regression classifier
are used to detect the normal or abnormal condition of the liver. The classifier
performance are analyzed by the bench mark parameters Sensitivity, Specificity, Accuracy,
Precision, Error Rate, Mathew Correlation Coefficient (MCC), and Classifier Success
Index (CSI) and compared. The Logistic regression achieved a higher accuracy of 80.95%
and outperformed other four classifiers.