Characterization of trabecular architecture in human femur radiographic images using directional multiresolution transform and adaboost model

Authors:
Thomas Christy Bobby;Swaminathan Ramakrishnan
Affiliations:
Department of Instrumentation Engineering, MIT Campus, Anna University, Chennai, India;Biomedical Engineering Group, Department of Applied Mechanics, IIT Madras, India
Venue:
SEMCCO'12 Proceedings of the Third international conference on Swarm, Evolutionary, and Memetic Computing
Year:
2012

Citing 4
Cited 0

An introduction to ROC analysis

Pattern Recognition Letters - Special issue: ROC analysis in pattern recognition
Assessment and Classification of Mechanical Strength Components of Human Femur Trabecular Bone Using Texture Analysis and Neural Network

Journal of Medical Systems
Using SOM and PCA for analysing and interpreting data from a P-removal SBR

Engineering Applications of Artificial Intelligence
The curvelet transform for image denoising

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this work, directional multiresolution curvelet transform is performed in radiographic images to characterize the trabecular structure. The trabecular regions of normal and abnormal human femur bone images are used for the study. The regions of interest such as femoral neck and head are analyzed and compared. The curvelet coefficients are calculated based on each scale and orientation for trabecular images. The mean and energy of the curvelet coefficients associated with each subband are computed. These values are used as the texture feature vector elements to evaluate changes taking place in the trabecular architecture. The three most significant mean and energy feature vector are found using principal component analysis and these values are used as an input to the Adaboost classifier. The results show that the architectural variations are more in the femoral neck when compared to femoral head. AdaBoost classifier performs better in terms of sensitivity (90%) and specificity (100%) for the chosen parameters for femoral neck region when compared to head regions.