A Novel Technique for Brain Tumour Detection Using IMADO

Malignant intracranial neoplasms are fatal with a diabolical prognosis that is deteriorating as the years go by because patients are diagnosed at a later stage. It is, hence, essential to perfect the segmentation of brain tumours to enhance proper treatment planning. Manual segmentation of radiology images is tiresome, and it becomes problematic to diagnose them. Artificial intelligence techniques have enhanced the accuracy and time taken to perform brain tumour segmentation, enhancing the treatment impacts. The Proposed Technique (IMADO) Improved Mean Absolute Deviation-Based Otsu technique addressed the issue. The (MADO) Mean Absolute Deviation-Based Otsu technique image thresholding algorithm overcomes the limitation by improving threshold value selection for segmenting images into front and background regions. The (IMADO) Improved Mean Absolute Deviation-Based Otsu technique is a promising methodology that substantially reduces time by incorporating heuristic methods for manual annotation, which benefits neurooncological clinical studies. This study uses anisotropic diffusion as a preprocessing step to minimize the noise in MRI images while keeping critical tumour boundaries, improving the IMADO's accuracy. Extensive testing has shown that the (IMADO) outperforms the original Otsu thresholding algorithm, resulting in more accurate and efficient brain tumour segmentation. The brain MRI images with unimodal or insignificant bimodal characteristic histograms can be segmented efficiently and accurately, and IMADO achieves a 4.62% accuracy improvement over Otsu's Thresholding Technique.