In medical image analysis, brain magnetic resonance image segmentation is still a difficult task. Therefore, it is crucial to provide an efficient method for identifying neurological illnesses and organising their therapies. In this work, we provide an approach to improve the segmentation of MRI images. To eliminate skull artefacts, morphological operators are first used for preprocessing. The first cluster centroids are then found by combining the Gaussian Mixture Model with K-means clustering. For better segmentation, we further integrate this with our Spatial Interval-Valued Intuitionistic Fuzzy C-Means clustering approach by adding
the Dα operator. This method effectively handles geographical information and uncertainty, improving segmentation performance. Existing approaches, including interval-valued intuitionistic fuzzy C-means clustering algorithm is contrasted with the suggested approach. Results from experiments show that our hybrid approach improves segmentation results.  The quantitative measures, the dice coefficient and Jaccard similarity, which are displayed in tables, attest to the fact that this method solves segmentation issues. and yields more trustworthy outcomes for clinical applications, making it a useful tool for medical diagnostics.