- Book Chapter
- 10.1007/978-3-031-83188-1_2
Semantic-Guided Feature Distillation for Multimodal Recommendation
- Jan 01, 2025
- Fan Liu + 2 more +2
Publications from 2021 to 2026
Showing 4 of 4 papers
Semantic-Guided Feature Distillation for Multimodal Recommendation
Dynamic Multimodal Fusion via Meta-Learning Towards Multimodal Recommendation
Electric impedance tomography enables portable and non-invasive approach to screen and monitor chronic kidney disease.
Chronic kidney disease (CKD) is an escalating global health concern, and non-invasive means for early CKD detection is eagerly awaited. Here, we explore the potential of using home-based frequency-difference electrical impedance tomography (fdEIT) to evaluate CKD based on bio-conductivity characteristics. We performed bio-conductivity measurement in vivo paired with standard estimated glomerular filtration rate (eGFR) measurements on a N=126 CKD patients by EIT and traditional blood and urine tests, respectively. We developed an EIT processing pipeline that extracts the kidney regions from EIT images. We further developed a regression model and a CKD classification scheme. Our results showed a significant correlation between EIT-features and eGFR, and the classification scheme shows sensitivity and specificity of 76.2% and 74.6% respectively considering stages 1 and 2 CKD versus stages 3, 4 and 5 CKD. These results suggest the feasibility of EIT to be used as a portable, self-administrated and home-based approach for CKD early diagnostic screening and longitudinal monitoring.Clinical Relevance-The results presented here demonstrates a cost-effective, home-based and self-administrative screening process on chronic kidney disease patients, thereby enhancing the quality and area of possible application of telemedicine. By achieving this, the process presented here can relieve the burden of public health system.
Read moreClinical validity of a negative venogram in patients with clinically suspected venous thrombosis.
Although it is generally accepted that negative venography excludes deep vein thrombosis (DVT) in patients in whom it is clinically suspected, there is no evidence to support this conclusion. To test the correctness of withholding anticoagulant therapy in these patients, we followed 160 consecutive patients who had clinically suspected DVT and negative venograms to determine the frequency of postvenographic DVT. Anticoagulant therapy was withheld in all patients. No patient died or developed pulmonary embolism during 3 months of follow-up. Two of the 160 patients (1.3%) attended the clinic on an emergency basis during follow-up with new symptoms of DVT and in both patients, DVT was confirmed by objective testing. These events developed within 5 days of venography, which suggests that they were induced by venography. Nevertheless, the findings indicate it is safe to withhold treatment in patients with clinically suspected DVT and negative venograms.
Read more