A5108023847- Multimodal Machine Learning Applications
- Bacterial Identification and Susceptibility Testing
- Cardiac Imaging and Diagnostics
- Venous Thromboembolism Diagnosis and Management
- Sepsis Diagnosis and Treatment
- Antibiotics Pharmacokinetics and Efficacy
- Trauma, Hemostasis, Coagulopathy, Resuscitation
- Antimicrobial Resistance in Staphylococcus
- Acute Myocardial Infarction Research
- Topic Modeling
- Video Analysis and Summarization
- Hemodynamic Monitoring and Therapy
- Advanced Causal Inference Techniques
Beijing Jishuitan Hospital
2022-2024
Peking University
2022-2024
Capital Medical University
2024
In this paper, we propose a prior guided transformer for accurate radiology reports generation. the encoder part, radiograph is firstly represented by set of patch features, which obtained through convolutional neural network and traditional encoder. Then an Additive Gaussian model applied to represent knowledge based on unsupervised clustering sparse attention. decoder embeddings are acquired probabilistically sampling from prior. visual language embeddings, fused our proposed Prior Guided...
Acute Myocardial Infarction (AMI) is a serve disease with rapid onset, high mortality and complex symptoms. Therefore, the accurate diagnosis effective treatment of AMI are important but difficult in clinic. In addition to symptoms electrocardiogram (ECG) features, laboratory detection echocardiogram (ECHO) provide significant help for physicians. With support big data, quite studies utilized machine learning investigate association between detection, ECHO onset AMI. However, most researches...