Graphical User Interface Based Computer Aided Diagnosis Tool of Human Brain Tumor Segmentation Through MRI and Validation

摘要

An important prerequisite for clinical analysis and treatment is a stage for medical image handling strategy that is more adaptable and accurate. A multifunctional graphical user interface (GUI) tool for computer aided diagnosis (CAD) that performs interactive image processing of brain tumor MRI images is presented in the proposed work. The proposed technique consists of different stages such as analysis, segmentation, Evaluation, Quantification, and Validation. The various functions implemented in the tool such as histogram equalization, Thresholding, image smoothing and clustering based segmentation, tumor detection along with some basic functions are explained in detail. In addition to this, the performance measures defined as probability random index (PRI), global consistency error (GCE), structural similarity (SSIM), variation of information (VOI) integrated in GUI-CAD is explained. Building effective Computer-Aided Detection and Diagnosis (CAD) systems involves the combination of running experiments, image mark up, security, analysis, evaluation, and validation in order to capture and evaluate medical images effectively. The system design enables radiologists to upload their feature sets and quickly compare the effectiveness of their methods against other stored feature sets.