Journal of Pure and Applied Ultrasonics

Open Access Open Access  Restricted Access Subscription Access

Image Classification Based on Textural Analysis of Ultrasound Images of Normal Human Liver


Affiliations
1 Pailan College of Management & Technology, West Bengal University of Technology, Kolkata-69, India
2 Microelectronics Division, Saha Institute of Nuclear Physics, 1/AF Bidhannagar, Kolkata-64, India
3 Department of Radiology, Chittaranjan National Cancer Institute, Kolkata-26, India
 

Texture analysis based on spatial gray-level dependence (SGLD) matrix computation was carried out over the input ultrasound images of livers to extract features, namely, maximum probability, uniformity, entropy, element difference moment of order 2, inverse element difference moment of order 2, homogeneity, correlation and contrast. In this work thirty ultrasound scan images of human normal livers as identified by the physician were collected from hospital. From each image portion of the liver, five sub-images were suitably cropped and taken as an input image for analysis. Finally, an unsupervised neural network learning technique viz. Self Organising Map (SOM) was used to classify normal liver parenchyma.
User
Notifications
Font Size

Abstract Views: 232

PDF Views: 0




  • Image Classification Based on Textural Analysis of Ultrasound Images of Normal Human Liver

Abstract Views: 232  |  PDF Views: 0

Authors

Anik Kumar Chakravorty
Pailan College of Management & Technology, West Bengal University of Technology, Kolkata-69, India
Subhasree Mukherjee
Pailan College of Management & Technology, West Bengal University of Technology, Kolkata-69, India
Kuntal Ghosh
Microelectronics Division, Saha Institute of Nuclear Physics, 1/AF Bidhannagar, Kolkata-64, India
Ratan Kumar Saha
Microelectronics Division, Saha Institute of Nuclear Physics, 1/AF Bidhannagar, Kolkata-64, India
Madhusudan Roy
Department of Radiology, Chittaranjan National Cancer Institute, Kolkata-26, India
Suparna Mazumdar
Microelectronics Division, Saha Institute of Nuclear Physics, 1/AF Bidhannagar, Kolkata-64, India

Abstract


Texture analysis based on spatial gray-level dependence (SGLD) matrix computation was carried out over the input ultrasound images of livers to extract features, namely, maximum probability, uniformity, entropy, element difference moment of order 2, inverse element difference moment of order 2, homogeneity, correlation and contrast. In this work thirty ultrasound scan images of human normal livers as identified by the physician were collected from hospital. From each image portion of the liver, five sub-images were suitably cropped and taken as an input image for analysis. Finally, an unsupervised neural network learning technique viz. Self Organising Map (SOM) was used to classify normal liver parenchyma.