Clustering analysis of feeg images using fuzzy c-means and k-means algorithms

This study presents an unsupervised clustering method for the segmentation and analysis of flat EEG (fEEG) images. Clustering identifies and groups distinct patterns or structural features within image data. The fEEG images are categorized using two established clustering methods: k-means for hard clustering and fuzzy c-means (FCM) for soft clustering. Hard clustering assigns each data point to a single cluster, whereas soft clustering clusters pixels into groups based on similarity. The k-means algorithm provides rapid segmentation with crisp boundaries. Meanwhile, FCM captures transitional zones and ambiguity through degrees of membership. Comparative results highlight the strengths of each technique, demonstrating the superior sensitivity of FCM in uncertain regions and the computational efficiency of K-Means.