A Survey on Bone Fracture Identification Techniques using Quantitative and Learning Based Algorithms
A. Saranya, K. Kottilingam
Abstract
Machine learning algorithms help to deeply analyses and monitor the minor changes in the objects. In the medical image, a little doubtful value also creates a big transformation in the outcome, which will be sorted out using learning-based algorithms. In this discussion, a genetic bone disorder Fibrous Dysplasia (FD) is considered, which replaces the bone tissues with fibrous tissues which lead to bone fracture or bone deformity. In this paper, three different bone investigation approaches are taken for discussion as Bone Mineral Density (BMD), Quantitative Measures for bone for analyzing thickness and, Algorithm based models are used for abnormality detections. The BMD test is used to calculate the life span of the bone and estimated fractural risk. The second analysis begins with quantitative measurements on different imaging techniques, which are applied to bone scan images to monitor the abnormality. The important imaging techniques of Quantitative Computer Tomography, High Resolution based Peripheral QCT, Quantitative Ultra Sound and other techniques are discussed. In model-based techniques, which automatically extracts the essential features from the image and groups them based on similar behavior, and discover the abnormal features from the image set.