ViSERA: Visualized & Standardized Environment for Radiomics Analysis - A Shareable, Executable, and Reproducible Workflow Generator

Medical image computing software development is essential for discovering novel image biomarkers in clinical research, which can improve patient outcomes by enabling more effective diagnosis and treatment strategies. However, a lack of standardized and user-friendly software environments has impeded their widespread adoption. This study presents the Visualized & Standardized Environment for Radiomics Analysis (ViSERA) software to address this challenge. The modular software provides a flexible platform for reliable and easy-to-use image reading from different modalities and formats, image registration, image fusion, image processing, image visualization, image segmentation, radiomics feature extraction, and machine learning model development. ViSERA offers a modular approach to medical image analysis, allowing users with different expertise levels to execute each module independently or combine them with other components. The output of one module can be fed as input to another module, enabling users to create custom, executable, and reproducible workflows with multiple steps. Further, the results of each module execution can be viewed directly in the visualization window, providing immediate feedback on the accuracy and effectiveness of each step. This modular approach offers flexibility and customization options to researchers and clinicians, making it easier to perform reliable and reproducible radiomics research in clinical practice settings. Furthermore, ViSERA offers a convenient way to save and share customized workflows with other users, promoting reproducibility, reusability, and transparency in medical image analysis. This feature is particularly useful for collaborative research projects and for ensuring consistency across different studies. This key feature and user-friendliness make it accessible for researchers at different skill levels, including radiation oncologists, radiologists, physicists, and data scientists.