مشخصات پژوهش

Segmentation of brain MRI images is a critical task in medical image analysis, yet existing clustering methods often struggle to produce accurate results due to various challenges. These include high computational complexity involved in computing distances between cluster centers and all pixels in each iteration, sensitivity to initial parameters and noise, and insufficient consideration of local spatial structures. This paper presents an innovative approach titled " Efficient Superpixel-Based Brain MRI Segmentation using Multi-Scale Morphological Gradient Reconstruction and Quantum Clustering" to address these challenges. The primary motivation is to develop an efficient and robust segmentation method capable of delivering accurate results while overcoming computational burdens and parameter sensitivity. To achieve this, we propose a multi-scale morphological gradient reconstruction operation to create precise superpixel images, enhancing the representation of local spatial structures. Leveraging these superpixel images, we efficiently compute histograms, effectively condensing the original color image information. Subsequently, quantum clustering is applied to these superpixel images using histogram parameters to achieve the desired segmentation outcome. Our experimental evaluation on brain MRI images demonstrates the superiority of the proposed method in terms of segmentation accuracy and processing speed compared to state-of-the-art clustering techniques. The results highlight the effectiveness of our approach in addressing the limitations of traditional methods, offering a promising solution for brain MRI segmentation in medical imaging applications.