Analysis of Medical Images using Image Registration Feature-based Segmentation Techniques

• Published in: 2022 2nd International Conference on Technological Advancements in Computational Sciences (ICTACS) pp. 485 - 490
• Main Authors: G, Sindhu Madhuri, K, Shashikala H.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 10.10.2022
• Subjects: Bones; Clustering; Error analysis; Feature extraction; Image registration; Image segmentation; Medical; Object segmentation; Partitioning algorithms; Registration; Segmentation
• DOI: 10.1109/ICTACS56270.2022.9987895

Image Segmentation is one of the very important optimistic and emerging fields in all image processing applications. It has a wide range of applications like machine vision, fingerprint recognition, digital forensics, medical imaging, and face recognition and so on. Based on specific application, various image segmentation techniques like thresholding, region growing, watershed, clustering algorithms, fuzzy algorithms etc., are used to segment or partition the input images, labels each pixel in the images, locate the points, edges, boundaries and objects to identify various problems in the medical images. Also the identification of important parameters, detection of fractures and diseases, to decrease the death rate of patients suffering from various health problems is challenging research work in medical images. In this paper, author carryout the analysis for the automatic detection of bone fracture in early stage by taking two input x-ray medical images that are captured at different timings. This process is carried out and registered in 4 stages: In first stage-acquire input images and perform pre-processing by using geometrical transformation and register the input images, in second stage- the registered image is segmented using adaptive k-means clustering method, in third stage- automatic detection of the important features in x-ray image is extracted using image registration feature-based method. Automatic feature extraction is carried out for the observation of bone fracture in initial phase to increase the complexity of geometrical alignments of input images. Finally in the fourth stage, the performance of the results is analyzed with respect to accuracy and error rate.