Virtual Booth

Accurate medical image segmentation is essential for identifying anatomical structures and pathological regions, yet its performance is often compromised by intensity inhomogeneity, which introduces non-uniform brightness variations and obscures tissue boundaries. To address this challenge, MedSIS is developed as a mathematical model-based medical image segmentation system incorporating the Selective Additive Bias Correction (SABC) model within a user-friendly graphical user interface (GUI). Based on a variational formulation, the proposed system simultaneously performs selective segmentation and bias field estimation, enabling accurate delineation of user-defined regions of interest under non-uniform intensity conditions. A marker-based initialization technique further facilitates targeted segmentation while minimizing interference from surrounding structures. The system is designed to support the analysis of medical images containing pathological regions, such as tumours and lesions, as well as anatomical structures. Experimental evaluations demonstrate promising segmentation performance in terms of Dice Similarity Coefficient (DSC), Jaccard Similarity Coefficient (JSC), and computational efficiency, highlighting the effectiveness of the SABC model in addressing intensity inhomogeneity. Beyond its technical capabilities, MedSIS offers a practical and interpretable platform with potential applications in computer-aided diagnosis (CAD), medical image analysis, and clinical research. Its GUI-based implementation also provides opportunities for future software integration and commercialization within healthcare and medical imaging environments.