Brain Tumor:
A brain tumor occurs when infrequent cells shape appears inside the cerebrum, which has two main types, namely, benign tumors and malignant tumors. Malignant tumors can be categorized into basic tumors, and secondary tumors that spread elsewhere. Automated brain tumor detection through MRI can offer a valued outlook and earlier accurate detection of the brain tumor. The tumor detection is performed in several stages, namely, enhancement, segmentation, classification. Depending on the tumorous tissue origin and its behavior, it is classified from less aggressive (benign: grades I and II) to aggressive (malignant: grades III and IV). The clinicians diagnose brain tumors using medical imaging techniques like magnetic resonance imaging(MRI) and computerized tomography (CT).
Magnetic Resonance Imaging:
Magnetic resonance imaging or MRI scanning uses magnetism, radio waves, and a computer to produce images of body structures. MRI imaging gives structural details of the brain and gives excellent results even during the early onset of disease. images. In general, the artifacts present in the MRI are partial volume, RF noise, intensity overlaps, motion, and gradient. The different MRI modalities are shown below. There are numerous challenges affecting automated or semiautomated brain tumor localization in MRI. The challenges are as follows.
- The tumorous tissues may originate anywhere in the brain; therefore, their localization is very challenging.
- All normal brain tissues have probable and prefixed locations. The tumor growth compresses normal tissues. deforming shape and making their identification even more difficult.
- In MRI, image sequences like benigs(T1-weighted, T2-weighted) malignant(T1c, fluid-attenuated inversion recovery(FLAIR) )are generated. Each sequence gives different biological information about brain tissues. So using only some of these sequences may not guarantee accurate tumor detection.
- The accuracy of segmentation also affects growth rate prediction, which is a very important issue for post-treatment planning.
Top comments (0)