<p>Internet of Medical Things (IoMT) is an effective network that has achieved significant progress of upcoming technology for heterogeneous and broadly associated networks. In medical applications, the network consists of different process for monitoring, diagnosing and treating the patients with various types of devices to attain the goal. An IoMT scheme in smart healthcare system observes the patients and the important signs of the normal and abnormal cases that can be detected according to the gathered data. Although, there are different clustering techniques, still, the performance of optimization in clustering techniques is an important research theory. To overcome, this paper aims to design the periodical discovering, managing, clustering, and analyzing useful data regarding potential patients using hybrid optimization-based unsupervised clustering. Here, the novel hybrid meta-heuristic algorithm is utilized for the clustering, in which the cluster centre node with information is selected. The data is grouped on the basis of distance and data characters. The optimum values are attained according to the multi-objective function with factors like accuracy of clustering, latency, and computational cost. Two well-performing algorithms are combined to design the hybrid Deviation-based Dingo-COA (DD-COA) algorithm. Consequently, after analyzing various components by solving different IoMT datasets, the capacity and the supremacy of the suggested model are well determined among its traditional approaches. From the findings, the proposed model demonstrates that it achieves the desired result for effectively handling the medical data through clustering approach in terms of 97.4% for accuracy and 96.8% for precision.</p>