With the expansion of healthcare and nursing data scales, coupled with the advancements in artificial intelligence techno-logies, the feature extraction techniques have become increasingly integral to the field of nursing, acting as a pivotal catalyst for intelligent transformation. This article provides a comprehensive review of multimodal feature extraction techniques and their application progress in areas such as early warning systems for high-risk nursing events, chronic disease management, adverse event management and wearable devices in nursing. The aim is to furnish a reference for the precise development and clinical application of nursing models.