School of Electronics Engineering, VIT-AP University, Amaravathi, AP, 522237, India.
School of Electronics Engineering, Vellore Institute of Technology, Vellore, Tamilnadu, 632014, India.
Sci Rep. 2024 Nov 30;14(1):29804. doi: 10.1038/s41598-024-81648-9.
The human-brain is a vital and complicated organ within the body. Identifying brain-related diseases can be challenging. Typically, Magnetic Resonance Imaging (MRI) scanning methods are used to gain insights of the protected regions in the body. Brain segmentation can result in identifying region boundaries as a set of contours. However, segmenting brain images poses several challenges, including noise, bias field, and partial volume effect (PVE). Removing noise, accurately segmenting tissues and tumors are crucial for effective evaluation. To enhance tissue and tumor segmentation, a new machine learning-based method called as Gaussian-Kernelized Enhanced Intuitionistic Fuzzy-C-Means (GKEIFCM) has been proposed. Approach enhances Improved Intuitionistic Fuzzy-C-Means Algorithm (IIFCM) by utilizing Gaussian kernelized distance between pixels, resulting in uncomplicated segmentation with reduced computational times and improved efficiency. This proposed novel method proved to be expertise in tissue and tumor classification and identification respectively. The results demonstrate the effectiveness of GKEIFCM interms of Dice, Jaccard-similarity-index, Accuracy and Execution time.
人脑是体内至关重要且复杂的器官。识别与大脑相关的疾病具有一定挑战性。通常，磁共振成像 (MRI) 扫描方法可用于深入了解身体的受保护区域。脑分割可将区域边界识别为一组轮廓。然而，脑图像分割存在若干挑战，包括噪声、偏置场和部分体积效应 (PVE)。去除噪声、准确分割组织和肿瘤对于有效评估至关重要。为了增强组织和肿瘤分割，提出了一种称为基于高斯核增强直觉模糊 C 均值 (GKEIFCM) 的新机器学习方法。该方法通过利用像素之间的高斯核化距离来增强改进的直觉模糊 C 均值算法 (IIFCM)，从而实现简化的分割，减少计算时间并提高效率。该新方法在组织和肿瘤分类和识别方面表现出色。结果表明，GKEIFCM 在骰子系数、Jaccard 相似性指数、准确性和执行时间方面具有有效性。
