Fuzzy C-Means (FCM) is a common clustering algorithm, and scholars usually use Multiple Population Genetic Algorithm (MPGA) to optimize the clustering centers. Still, MPGA has insufficient global search capabilities and lacks selfadaptability, so the optimized clustering centers are not optimal and eventually tend to converge prematurely. Therefore, this paper proposes an adaptive FCM clustering algorithm, referred to as DMGA-FCM, based on Derived Multi-population Genetic Algorithm (DMGA). The first proposed derivation operator for DMGA-FCM considers the problem of insufficient optimization ability among populations. This operator performs derivative operations on the initial population to improve the algorithm's optimization-seeking capability. At the same time, the adaptive probability fuzzy control operator dynamically adjusts the genetic probability to improve the algorithm adaptability, thereby enhancing the global optimization capability of the DMGA algorithm and avoiding premature convergence problems. Finally, this is integrated into the FCM algorithm. The analysis of simulation experiments and MRI brain map application results shows that DMGA-FCM outperforms other competitive methods in medical imaging segmentation and clustering.