Long Bone Fracture Detection Using Machine Learning


  • Najat Abduh Hezam Nagi Al-Shara'abi
  • Nashwan Ameen Abdulwahab Al-khulaidi


Long Bone Fraction Detection, Normal, Abnormal, Classifiers, Machine Learning.


In this work, the researcher presents an automatic system to detect the presence of long bone fractures by using clinical images obtained from X-Ray. The procedure for the diagnosis of the bone fractures is considered to be a very critical step based on factors to identify this image as normal or abnormal to save the effort and time spent to detect bone fractures. Trained radiologists often identify rare diseases with high accuracy such as fractures. Accurate diagnosis of the bone fraction is important. The Histogram Oriented Gradients (HOG) and Local Binary Pattern Algorithm (LBP) are used for features extraction. This study used two different classifiers. The first classification is Support Vector Machine (SVM), which provides accuracy of 97.85 percent by Radial basis kernel function (RBF) and the second classifier is Multilayer Perceptron (MLP), which gives accuracy of 99.15 percent, then the accuracy of the classifiers are compared with each other. Consequently, Multilayer Perceptron algorithm has the highest accuracy of 99.15 percent. We obtained the best results by MLP using LBP which has the best results as Sensitivity, Specificity and Accuracy are 100, 98.35 and 99.15 percent. The study presents a discussion and discovery of a computer-based long bone fracture detection system by MATLAB .The purpose of this work is to provide insight into the related activities of research conducted. In addition, the researcher proposed long bone fraction detection system by using a computer-supported program.




How to Cite

Najat Abduh Hezam Nagi Al-Shara’abi, & Nashwan Ameen Abdulwahab Al-khulaidi. (2022). Long Bone Fracture Detection Using Machine Learning. Progress in Engineering Application and Technology, 3(2), 998–1008. Retrieved from https://publisher.uthm.edu.my/periodicals/index.php/peat/article/view/9815