Publications by authors named "FengHua Cai"
Background: Pulse oximeters are vital for assessing blood oxygen levels but can produce inaccurate readings during patient motion, leading to false alarms and alarm fatigue. Analog SpO Technology, which uses analog waveforms to filter motion artifacts, may improve accuracy compared to digital sensors. However, the effectiveness of this technology in reducing false alarms in clinical settings remains unclear.
View Article and Find Full Text PDFArticle Synopsis
- Venous thromboembolism (VTE) is a serious medical condition that is difficult to treat, and creating a knowledge graph (VTEKG) can help organize and analyze related information better, but current methods have issues with errors and redundancy.
- The study introduces a new deep learning model called BCSLinker aimed at improving the construction of VTEKG by effectively extracting relevant medical entities and their relationships while reducing errors and redundant information.
- The model was tested using data from VTE patients, achieving an impressive F1 score of 86.9%, outperforming existing models and contributing to the development of a helpful question-answering system for VTE-related inquiries.