dtComb: A Comprehensive R Library and Web Tool for Combining Diagnostic Tests

The combination of diagnostic tests has become a crucial area of research, aiming to improve the accuracy and robustness of medical diagnostics. While existing tools focus primarily on linear combination methods, there is a lack of comprehensive tools that integrate diverse methodologies. In this study, we present dtComb, a comprehensive R package and web tool designed to address the limitations of existing diagnostic test combination platforms. One of the unique contributions of dtComb is offering a range of 142 methods to combine two diagnostic tests, including linear, non-linear, machine learning algorithms, and mathematical operators. Another significant contribution of dtComb is its inclusion of advanced tools for ROC analysis, diagnostic performance metrics, and visual outputs such as sensitivity-specificity curves. Furthermore, dtComb offers classification functions for new observations, making it an easy-to-use tool for clinicians and researchers. The web-based version is also available for non-R users, providing an intuitive interface for test combination and model training.