Federated learning (FL) is a promising approach for healthcare informatics that enables the training of a shared global model without sharing sensitive patient data. This method allows multiple institutions to collaborate on data-driven insights while preserving privacy. The paper reviews FL technologies, focusing on challenges such as statistical heterogeneity, communication efficiency, and privacy issues in the healthcare context. It discusses general solutions to these challenges and highlights the potential of FL in healthcare applications, including predictive modeling, patient similarity learning, and phenotyping. The paper also addresses the need for improved data quality, integration of expert knowledge, incentive mechanisms, personalization, model precision, and funding for FL research. FL offers a way to analyze fragmented healthcare data without compromising patient privacy, making it a valuable tool for improving healthcare outcomes. The review emphasizes the importance of addressing open questions and challenges in FL for healthcare, including data standardization, model personalization, and efficient communication protocols. The paper concludes that FL has significant potential to advance healthcare informatics by enabling collaborative, privacy-preserving data analysis across multiple institutions.Federated learning (FL) is a promising approach for healthcare informatics that enables the training of a shared global model without sharing sensitive patient data. This method allows multiple institutions to collaborate on data-driven insights while preserving privacy. The paper reviews FL technologies, focusing on challenges such as statistical heterogeneity, communication efficiency, and privacy issues in the healthcare context. It discusses general solutions to these challenges and highlights the potential of FL in healthcare applications, including predictive modeling, patient similarity learning, and phenotyping. The paper also addresses the need for improved data quality, integration of expert knowledge, incentive mechanisms, personalization, model precision, and funding for FL research. FL offers a way to analyze fragmented healthcare data without compromising patient privacy, making it a valuable tool for improving healthcare outcomes. The review emphasizes the importance of addressing open questions and challenges in FL for healthcare, including data standardization, model personalization, and efficient communication protocols. The paper concludes that FL has significant potential to advance healthcare informatics by enabling collaborative, privacy-preserving data analysis across multiple institutions.