A5100319043- Topic Modeling
- Advanced Graph Neural Networks
- Machine Learning in Healthcare
- Data Quality and Management
- Security and Verification in Computing
- Ophthalmology and Eye Disorders
- Advanced Memory and Neural Computing
- Facial Rejuvenation and Surgery Techniques
- Cardiac Valve Diseases and Treatments
- Physical Unclonable Functions (PUFs) and Hardware Security
- Renal Transplantation Outcomes and Treatments
- Advanced Computational Techniques and Applications
- Cryptographic Implementations and Security
- Artificial Intelligence in Healthcare
- Hematopoietic Stem Cell Transplantation
- Atrial Fibrillation Management and Outcomes
- Cardiac tumors and thrombi
- Computational Physics and Python Applications
- Cardiac, Anesthesia and Surgical Outcomes
- Cerebral Venous Sinus Thrombosis
- Sentiment Analysis and Opinion Mining
- Smart Grid Security and Resilience
- Cardiac Imaging and Diagnostics
- Pericarditis and Cardiac Tamponade
- Neural Networks Stability and Synchronization
Cloud Computing Center
2019-2025
New York University
2024
Pudong Medical Center
2024
Fudan University
2024
Zhongshan Hospital
2024
Guangzhou Medical University Cancer Hospital
2024
Guangzhou Medical University
2024
First People's Hospital of Yunnan Province
2022-2023
Xi'an Jiaotong University
2021-2023
Kunming University of Science and Technology
2022-2023
Clustering is a widely used technique in data mining applications to discover patterns the underlying data. Most traditional clustering algorithms are limited handling datasets that contain either continuous or categorical attributes. However, with mixed types of attributes common real life problems. In this paper, we propose distance measure enables both and This derived from probabilistic model between two clusters equivalent decrease log-likelihood function as result merging. Calculation...
This paper proposes a new memory controller design that enables secure sharing of main among mutually mistrusting parties by eliminating timing channels. study demonstrates shared controllers are vulnerable to both side channel and covert attacks exploit interference as To address this vulnerability, we identify the sources in conventional design, propose protection scheme eliminate across security domains through two changes: (i) per domain based queueing structure, (ii) static allocation...
In today's multicore processors, the last-level cache is often shared by multiple concurrently running processes to make efficient use of hardware resources. However, previous studies have shown that a vulnerable timing channel attacks leak confidential information from one process another. Static partitioning can eliminate channels but incurs significant performance overhead. this paper, we propose Secure Dynamic Cache Partitioning (SecDCP), technique defeats attacks. The SecDCP scheme...
Information security can be compromised by leakage via low-level hardware features. One recently prominent example is cache probing attacks, which rely on timing channels created caches. We introduce a design language, SecVerilog, makes it possible to statically analyze information flow at the level. With systems built with verifiable control of and other channels. SecVerilog Verilog, extended expressive type annotations that enable precise reasoning about flow. It also comes rigorous formal...
Abstract Objective This study investigates speech decoding from neural signals captured by intracranial electrodes. Most prior works can only work with electrodes on a 2D grid (i.e., Electrocorticographic or ECoG array) and data single patient. We aim to design deep-learning model architecture that accommodate both surface (ECoG) depth (stereotactic EEG sEEG) The should allow training multiple participants large variability in electrode placements the trained perform well unseen during...
Currently, the significance of multimodal sentiment analysis is progressively increasing. However, heterogeneity signals poses a challenge in learning modality representation and fusing information. To address it, BiMSA (multimodal based on BiGRU bidirectional interactive attention) proposed this paper. In layer, incorporates structure consisting to extract contextual information from video audio inputs. Subsequently, features are projected into internal representations for extracting This...
Abstract Background Diabetic kidney disease (DKD) is a serious complication of diabetes mellitus (DM), with patients typically remaining asymptomatic until reaching an advanced stage. We aimed to develop and validate predictive model for DKD in initial diagnosis type 2 (T2DM) using real-world data. Methods retrospectively examined data from 3,291 (1740 men, 1551 women) newly diagnosed T2DM at Ningbo Municipal Hospital Traditional Chinese Medicine (2011–2023). The dataset was randomly divided...
Background: Early identification of patients at high risk postoperative acute kidney injury (AKI) can facilitate the development preventive approaches. This study aimed to develop prediction models for AKI in noncardiac surgery using machine learning algorithms. We also evaluated predictive performance that included only preoperative variables or important predictors. Materials and methods: Adult undergoing were retrospectively (76,457 discovery cohort 11,910 validation cohort). was...
Abstract Purpose Pediatric patients with inborn errors of immunity (IEI) undergoing umbilical cord blood transplantation (UCBT) are at risk early mortality. Our aim was to develop and validate a prediction model for mortality after UCBT in pediatric IEI based on pretransplant factors. Methods Data from 230 who received their first between 2014 2021 single center were analyzed retrospectively. 2014–2019 2020–2021 used as training validation sets, respectively. The primary outcome interest...
Purpose Teprotumumab, a specific blocking antibody to the insulin like growth factor 1 receptor, significantly reduced proptosis in patients with thyroid eye disease (TED) recent clinical trials. Given its specificity, we expect it demonstrate greater efficacy on worse affected orbit, asymmetric TED. Herein, investigate differential impact of teprotumumab orbits such patients. Methods In this pooled analysis who were enrolled phase 2 ( NCT01868997 ) and 3 NCT03298867 trials, all TED...
Background: Thyroid eye disease (TED) causes orbital soft-tissue expansion. Recent studies have suggested that brow and temple changes may also occur. Teprotumumab, a monoclonal antibody to the insulin-like growth factor 1 receptor reduces swelling in TED. In this study, we quantified pan facial volumes eyelid position, following treatment with teprotumumab. Methods: prospective consecutive patients who were treated teprotumumab appraised for study eligibility. All had 3D imaging using...
Information security can be compromised by leakage via low-level hardware features. One recently prominent example is cache probing attacks, which rely on timing channels created caches. We introduce a design language, SecVerilog, makes it possible to statically analyze information flow at the level. With systems built with verifiable control of and other channels. SecVerilog Verilog, extended expressive type annotations that enable precise reasoning about flow. It also comes rigorous formal...
Abstract In order to solve the cooperative search problem of multiple unmanned aerial vehicles (multi-UAVs) in a large-scale area, we propose genetic algorithm (GA) incorporating simulated annealing (SA) for solving task region allocation among multi-UAVs on premise that large area is divided into several small areas. Firstly, describe be solved, and regard areas as traveling salesman (MTSP). And objective function established under number searched by each UAV balanced. Then, improve GA,...
Abstract Forecasting of severe acute graft-versus-host disease (aGVHD) after transplantation is a challenging ‘large p , small n ’ problem that suffers from nonuniform data sampling. We propose dynamic probabilistic algorithm, daGOAT, accommodates sampling heterogeneity, integrates multidimensional clinical and continuously updates the daily risk score for aGVHD onset within two-week moving window. In studied cohorts, cross-validated area under receiver operator characteristic curve (AUROC)...
Background Congenital heart defects (CHDs) are the most common type of birth defects. The genetic aetiology CHD is complex and incompletely understood. overall distribution causes in patients with from neonatal intensive care units (NICUs) needs to be studied. Methods cases were extracted China Neonatal Genomes Project (2016–2021). Next-generation sequencing results medical records retrospectively evaluated note frequency diagnosis respective patient outcomes. Results In total, 1795...
Abstract Context The components of metabolic syndrome (MetS) are interrelated and associated with renal complications in patients type 2 diabetes (T2D). Objective We aimed to reveal prevalent profiles T2D identify which were risk markers for progression. Methods A total 3556 participants from a hospital (derivation cohort) 931 community survey (external validation included. primary outcome was the onset diabetic kidney disease (DKD), secondary outcomes included estimated glomerular...
Abstract Background Venous thromboembolism (VTE) risk assessment in surgical patients is important for the appropriate diagnosis and treatment of patients. The commonly used Caprini model limited by its inadequate ability to discriminate between stratums on population southwest China lengthy factors. purpose this study was establish an improved VTE that accurate simple. Methods This based clinical data from 81,505 hospitalized Southwest Hospital January 1, 2019 June 18, 2021. Among...
Abstract Background Childhood obesity has become a huge challenge to childhood health, and there is lack of understanding about the relationship between dietary inflammatory index (DII) obesity. The main objective this study was analyze potential link DII among children adolescents residing in United States. Methods A cross-sectional analysis performed using data obtained from National Health Nutrition Examination Survey 2009 2018. In total, 12,454 participants were included analysis....