Digitization of the healthcare industry is reshaping the lifestyle of the future generation. Augmentation of utterly supreme technologies like Machine Learning (ML), Deep Learning (DL), Artificial Intelligence (AI), Software-Defined Networking (SDN), Federated Learning (FL), Network Softwarization and virtualization, Quantum Computation, Augmented and Virtual Reality, etc. technologies changes the mode of traditional e-healthcare infrastructure. Most importantly data-driven AI is emerging as one of the most fundamental approaches for building robust and appropriate statistical models from healthcare-related data. Generally, the data collection inside e-healthcare systems is done from edge-based sensor devices or wireless sensor networks. This huge data needs space, privacy, security, reliable care, and prolonged restoration. Healthcare data is always sensible and demands ultimate caution while orchestration over-distributed entities. The existing ML techniques are not fully capable to restrict data privacy due to their stations in data silos. A clever approach to tackle such distributed ML learning securely is to introduce the FL techniques for big medical data and associated resource distributions. This study primarily focuses on this key fact and proposes a holistic framework for e-healthcare systems leveraging the aggregated FL approaches using Edge-Cloud interplay. We have further incorporated the premium solution approaches to the existing challenges in the world of digital healthcare system.