Publication Date: 2023/10/06
Abstract: The most anticipated procedures in the present digital era are image retrieval from vast libraries of health care sector data. The objective of this work is to use machine learning techniques to efficiently classify medical images from larger healthcare datasets. In this paper, features are retrieved using GLCM (Grey Level Co-occurrence Matrix) from five different classes of medical images using image properties including dissimilarity, correlation, homogeneity, contrast, ASM, and energy. To extract the attributes utilizing the layers, the photographs are looked at from a variety of angles (0, 45, 90, and 135). Input is made to the trendiest Machine Learning (ML) models using the received feature vectors, that is Random Forest (RF) and Support vector machines (SVM) are used to classify images. Performance evaluation is carried out by contrasting and assessing the experimental results of SVM and Random Forest, two algorithms produces classification accuracy of 95.22% and 96.37%, respectively. When picture classification is done on bigger healthcare datasets, the Precision, Recall, and F1-Score are compared as well, and they are more accurate. In comparison, the feature extraction employing GLCM and ML models can extract better medical image features and achieve higher classification accuracy.
Keywords: Grey Level Co-occurrence Matrix, Cloud Computing, Random Forest (RF) and Support Vector Machines (SVM), Content-Based Image Retrieval.
DOI: https://doi.org/10.5281/zenodo.8413591
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT23SEP1738.pdf
REFERENCES