Transmission of confidential information through communication channel faces several challenges of security. In medical field, sharing of medical images with specialist doctors are essentially required for treatment and diagnosis of diseases. So secured medical data transmission is vital for copyright and ownership protection. However, in recent digital technologies there are possibilities of several attacks on medical data during transmission. The digital medical image watermarking is usually used for transmission of secured medical data and copyright protection. In this paper, various types of medical images and its datasets, formats, watermark features and performance parameters of watermarking are investigated. Trade off between Watermark features like robustness, imperceptibility and capacity has studied. In addition, review of different techniques, classification, algorithm, and comparative study of watermarking system is presented. The conventional watermarking techniques using spatial and frequency domain are not capable to balance between watermarking features, so hybrid or novel technique using deep learning is solution for balancing features.