Medical imaging is essential for diagnosing illnesses, and due to the delicate nature of medical images, stringent security and privacy measures must be in place. Medical images should be secured before being outsourced in a cloud-based medical system for Healthcare Industry 4.0. However, it is difficult and currently impractical to process queries over encrypted data without first performing the decryption step. In the paper, we suggest an effective method for locating the precise nearest neighbour in a set of encrypted medical photos. By obtaining the lower bound of the Euclidean distance, which is correlated with the mean and standard deviation of the data, we can eliminate candidates instead of computing the Euclidean distance. Our method can find the precise nearest neighbour as opposed to an approximation, unlike the majority of other existing approaches. We then assess our suggested strategy to show its usefulness.