Automatic Detection of Brain Abnormalities and Tumor Segmentationin MRI Sequence

This paper presents an automated and clinicallytested method for detection of brain abnormalities and tumor-edema segmentation using the MRI sequences. It follows a Radiologist’s approach to the brain diagnosis using multiple MRI sequences instead of any prior models or training phases. Our procedure consists of the following steps: a) Pre-processing of the MRI sequences, T2, T1 and T1 post contrast for size standardization, contrast equalization and division into active cells b) Identification of the T2 MRI sequence as normal or abnormal by exploiting the vertical symmetry of the brain c) Determination of the region of abnormality using its hyper-intense nature. d) Separation of tumor from edema using the T1 and its postcontrast (enhanced) sequences and e) Estimation of the volume of tumor found and generation of an anatomical differential of the possible disorders. This approach has been tested on more than a hundred real dataset both normal and abnormal representing tumors of different shapes, locations and sizes and results being checked by radiologists thus defining an efficient and robust technique for automated detection and segmentation of brain tumors. Keywords-Automated brain tumor detection; Tumoredema segmentation; robust estimation; Brain symmetry; Intensity outliers; Brain anatomical classification; T2, T1 MR images