Improvement of Accuracy in Prevention of Medical Images from Security Threats Using Novel Lasso Regression in Comparison with K‐Means Classifier
K. R. Raghul, M Kalaiyarasi · 2025
This study focuses on enhancing the security of medical images by analyzing the effectiveness of two machine learning algorithms: novel lasso regression and k-means classifier. The primary objective is to compare their performance in detecting and preventing security threats in medical image datasets. A total of 20 samples were used, divided into two groups: Group 1 comprised of novel lasso regression (N=10), and Group 2 consisted of k-means classifier (N=10). The statistical power of the study, measured by the G*Power value, was set at 0.95 to ensure reliable results. The analysis revealed that the novel lasso regression model achieved an accuracy of 92.64%, significantly outperforming the k-means classifier, which recorded an accuracy of 88.47%. The statistical significance of these results was confirmed through an independent t-test, yielding a significance value of 0.037 (p<0.05), indicating that the difference in accuracy between the two models is statistically significant. This demonstrates that the novel lasso regression method is more effective in enhancing the security of medical images than the k-means classifier. The novel lasso regression algorithm offers a more robust and accurate solution for medical image security, making it a valuable tool for protecting sensitive patient information from potential data breaches and security threats.