Dan Lin

Dan Lin


Research Focus & Metrics

Representative Publications

View All Publications
1 December 2023 in Korean Journal of Anesthesiology
Verified
Journal Article
Open access
PDF file
LYLin Yang
Lin Yang
YCYongheng Cai
Yongheng Cai
DLDan Lin
Dan Lin
HHHe Huang
He Huang
BCBing Chen
Bing Chen
Background: Mechanical ventilation, particularly one-lung ventilation (OLV), can cause pulmonary dysfunction. This meta-analysis assessed the effects of dexmedetomidine on the pulmonary function of patients receiving OLV. Methods: The Embase, PubMed, MEDLINE, Cochrane Library, ClinicalTrials.gov, and Chinese Clinical Trial Registry databases were systematically searched. The primary outcome was oxygenation index (OI). Other outcomes including the incidence of postoperative complications were assessed. Results: Fourteen randomized controlled trials involving 845 patients were included in this meta-analysis. Dexmedetomidine improved the OI at 30 (mean difference [MD]: 40.49, 95% CI [10.21, 70.78]), 60 (MD: 60.86, 95% CI [35.81, 85.92]), and 90 min (MD: 55, 95% CI [34.89, 75.11]) after OLV and after surgery (MD: 28.98, 95% CI [17.94, 40.0]) and improved lung compliance 90 min after OLV (MD: 3.62, 95% CI [1.7, 5.53]). Additionally, dexmedetomidine reduced the incidence of postoperative pulmonary complications (odds ratio: 0.44, 95% CI [0.24, 0.82]) and length of hospital stay (MD: −0.99, 95% CI [−1.25, −0.73]); decreased tumor necrosis factor-α, interleukin (IL)-6, IL-8, and malondialdehyde levels; and increased superoxide dismutase levels. However, only the results for the OI and IL-6 levels were confirmed by the sensitivity and trial sequential analyses.Conclusions: Dexmedetomidine improves oxygenation in patients receiving OLV and may additionally decrease the incidence of postoperative pulmonary complications and shorten the length of hospital stay, which may be related to associated improvements in lung compliance, anti-inflammatory effects, and regulation of oxidative stress reactions. However, robust evidence is required to confirm these conclusions.

17 May 2023 in arXiv
Verified
Preprint
Open access
PDF file
KWKai Wang
Kai Wang
SLSiqiang Luo
Siqiang Luo
DLDan Lin
Dan Lin
We study Graph Neural Networks (GNNs)-based embedding techniques for knowledge graph (KG) reasoning. For the first time, we link the path redundancy issue in the state-of-the-art KG reasoning models based on path encoding and message passing to the transformation error in model training, which brings us new theoretical insights into KG reasoning, as well as high efficacy in practice. On the theoretical side, we analyze the entropy of transformation error in KG paths and point out query-specific redundant paths causing entropy increases. These findings guide us to maintain the shortest paths and remove redundant paths for minimized-entropy message passing. To achieve this goal, on the practical side, we propose an efficient Graph Percolation Process motivated by the percolation model in Fluid Mechanics, and design a lightweight GNN-based KG reasoning framework called Graph Percolation Embeddings (GraPE). GraPE outperforms previous state-of-the-art methods in both transductive and inductive reasoning tasks while requiring fewer training parameters and less inference time.

26 April 2023 in arXiv
Verified
Preprint
Open access
PDF file
XZXinliang Zhou
Xinliang Zhou
DLDan Lin
Dan Lin
ZJZiyu Jia
Ziyu Jia
Jiaping XiaoJiaping Xiao
Jiaping Xiao
CLChenyu Liu
Chenyu Liu
YLYang Liu
Yang Liu
Drowsy driving has a crucial influence on driving safety, creating an urgent demand for driver drowsiness detection. Electroencephalogram (EEG) signal can accurately reflect the mental fatigue state and thus has been widely studied in drowsiness monitoring. However, the raw EEG data is inherently noisy and redundant, which is neglected by existing works that just use single-channel EEG data or full-head channel EEG data for model training, resulting in limited performance of driver drowsiness detection. In this paper, we are the first to propose an Interpretability-guided Channel Selection (ICS) framework for the driver drowsiness detection task. Specifically, we design a two-stage training strategy to progressively select the key contributing channels with the guidance of interpretability. We first train a teacher network in the first stage using full-head channel EEG data. Then we apply the class activation mapping (CAM) to the trained teacher model to highlight the high-contributing EEG channels and further propose a channel voting scheme to select the top N contributing EEG channels. Finally, we train a student network with the selected channels of EEG data in the second stage for driver drowsiness detection. Experiments are designed on a public dataset, and the results demonstrate that our method is highly applicable and can significantly improve the performance of cross-subject driver drowsiness detection.