Privacy-Preserving Machine Learning for Healthcare: Open Challenges and Future Perspectives

Machine Learning (ML) has recently shown tremendous success in modeling various healthcare prediction tasks, ranging from disease diagnosis and prognosis to patient treatment. Due the sensitive nature of medical data, privacy must be considered along entire ML pipeline, model training inference. In this paper, we conduct a review recent literature concerning Privacy-Preserving (PPML) for healthcare. We primarily focus on privacy-preserving inference-as-a-service, perform comprehensive existing trends, identify challenges, discuss opportunities future research directions. The aim is guide development private efficient models healthcare, with prospects translating efforts into real-world settings.