Web based application system for cancerous object detection in ultrasound images

Computer Aided Diagnosis Systems-CAD is an interpretation support system for Radiologists as a second opinion in the diagnosis of cancer malignancy on medical imaging such as ultrasound. The latest CAD technology still detects the region of interest (RoI) of cancerous objects manually and is integrated into the ultrasound machine. In this study, a web-based application by adopting the MoRbAC method is proposed to automatically detect cancerous objects in ultrasound images for CAD development. MoRbAC is a combinatorial framework consisting of a simplified Chan-Vese (CV) active contour model and image morphological operations. Simplified CV serves to segment objects globally while morphological operations play a role in reducing detection errors and localizing them specifically. The proposed web-based application was validated on 20 ultrasound images of breast lesions and thyroid nodules. Python programming with Flask framework is utilized to deploy this application. The validation results are compatible across various browsers and achieve an average accuracy of 98.75%. This achievement shows that MoRbAC is feasible to be used as an ultrasound CAD detection technique in a web service system.