A5082553860- Colorectal Cancer Surgical Treatments
- Topic Modeling
- Gastric Cancer Management and Outcomes
- Enhanced Recovery After Surgery
- Natural Language Processing Techniques
- Text Readability and Simplification
- Statistical Methods in Clinical Trials
- Handwritten Text Recognition Techniques
- Image Processing and 3D Reconstruction
- Cardiac, Anesthesia and Surgical Outcomes
- Clinical Nutrition and Gastroenterology
- Health Systems, Economic Evaluations, Quality of Life
- Image Retrieval and Classification Techniques
- Text and Document Classification Technologies
- Multimodal Machine Learning Applications
- Optimal Experimental Design Methods
- Hand Gesture Recognition Systems
- Metastasis and carcinoma case studies
- Software Engineering Research
- Statistical Methods and Bayesian Inference
- Infrared Target Detection Methodologies
- Sentiment Analysis and Opinion Mining
- Human Pose and Action Recognition
- Advanced Causal Inference Techniques
- Esophageal and GI Pathology
Nanjing University
2014-2025
Anhui Medical University
2025
Second Hospital of Anhui Medical University
2025
Zhejiang University
2005-2025
Stomatology Hospital
2025
State Key Laboratory of Polymer Physics and Chemistry
2024
Chinese Academy of Sciences
2024
Changchun Institute of Applied Chemistry
2024
Shenzhen University
2024
Wuhan University of Science and Technology
2024
Propensity score matching is a method to reduce bias in non-randomized and observational studies. mainly applied two treatment groups rather than multiple groups, because some key issues affecting its application remain unsolved, such as the distance, assessment of balance baseline variables, choice optimal caliper width. The primary objective this study was compare propensity methods using different calipers choose width for use with three groups. authors used widths from 0.1 0.8 pooled...
To improve the performance of crop models for regional yield estimates, a particle filter (PF) was introduced to develop data assimilation strategy using Crop Environment Resource Synthesis (CERES)-Wheat model. Two experiments involving winter wheat estimations were conducted at field plot and on scale test feasibility PF-based analyze effects PF parameters spatiotemporal scales assimilating observations model assimilation. The significant improvements in estimation suggest that is feasible....
Emotion-cause pair extraction is a recently proposed task that aims at extracting all potential clause-level pairs of emotion and cause in text. To solve this task, researchers first two-step pipeline method. This method extracts the emotions causes individually step, then extracted filters invalid emotion-cause second step. Due to has error accumulation problem hard be optimized jointly, several one-step end-to-end models have been proposed. These share similar underlying idea, is,...
Emotion-cause pair extraction (ECPE) is a new task which aims at extracting the potential clause pairs of emotions and corresponding causes in document. To tackle this task, two-step method was proposed by previous study first extracted emotion clauses cause individually, then paired clauses, filtered out without causality. Different from that separated detection matching into two steps, we propose Symmetric Local Search Network (SLSN) model to perform simultaneously local search. SLSN...
Background The etiology of interstitial cystitis/painful bladder syndrome (IC/BPS) remains elusive, presenting significant challenges in both diagnosis and treatment. To address these challenges, we employed a comprehensive approach aimed at identifying diagnostic biomarkers that could facilitate the assessment immune status individuals with IC/BPS. Methods Transcriptome data from IC/BPS patients were sourced Gene Expression Omnibus (GEO) database. We identified differentially expressed...
Abstract Background Epidemiological studies have confirmed the relationship between personality trait neuroticism and physical health. However, frailty remains unconfirmed. This study employed a bi-directional two-sample Mendelian randomization (MR) approach to investigate causal frailty. Methods The genome-wide association (GWAS) data from UK Biobank contained twelve neuroticism-related traits with 489,212 participants. genetic index were extracted Swedish TwinGene, involving 175,226...
Zero-Shot Cross-Modal Retrieval (ZS-CMR) aims to perform cross-modal retrieval on data of unseen classes, where a key challenge is how address the modality-gap and domain-shift problems simultaneously. Existing methods tackle this mainly by embracing sample-label alignment paradigm, which aligns samples different modalities but same class with word embedding their label. However, these only focus class-level overlook rich fine-grained semantic information in samples, incurring coarse...
Emotion-cause pair extraction (ECPE) is a recently proposed task that aims to extract the potential clause pairs of emotions and its corresponding causes in document. In this article, we propose new paradigm for ECPE task. We cast as two-turn machine reading comprehension (MRC) task, i.e., transformed identifying answer clauses from input document specific query. This MRC formalization brings several key advantages: First, QA manner provides an explicit pairing way identify target emotion;...
In the field of portrait video generation, use single images to generate videos has become increasingly prevalent. A common approach involves leveraging generative models enhance adapters for controlled generation. However, control signals (e.g., text, audio, reference image, pose, depth map, etc.) can vary in strength. Among these, weaker conditions often struggle be effective due interference from stronger conditions, posing a challenge balancing these conditions. our work on we identified...
Gastric cancer represents a great challenge for healthcare providers and requires multidisciplinary treatment approach in which surgery plays major role. Minimally invasive has been progressively developed, first with the advent of laparoscopy recently spread robotic surgery, but number issues are currently being debated, including limitations performing an effective extended lymph node dissection, real advantages systems, role Advanced Cancer, reproducibility total intracorporeal technique...
Colon cancer remains one of the most prevalent cancers worldwide. Unfortunately, there are no recognized and effective therapeutic strategies to prevent tumor recurrence after radical resection chemotherapy, disease-free survival (DFS) in patients with stage IIIB or IIIC disease unsatisfactory. Xian-Lian-Jie-Du optimization decoction (XLJDOD) is a Chinese herbal medicine (CHM) empirical prescription, which has been validated experimentally clinically that could inhibit progression colorectal...
Recognizing emotions from multi-modal data is an emotion recognition task that requires strong representation ability. The general approach to this naturally train the model on training without intervention. However, such natural scheme prone modality bias of (i.e., tending over-encode some informative modalities while neglecting other modalities) and overfit data). These biases may lead instability (e.g., performing poorly when neglected dominant for recognition) weak generalization unseen...
The success of the text-guided diffusion model has inspired development and release numerous powerful models within open-source community. These are typically fine-tuned on various expert datasets, showcasing diverse denoising capabilities. Leveraging multiple high-quality to produce stronger generation ability is valuable, but not been extensively studied. Existing methods primarily adopt parameter merging strategies a new static model. However, they overlook fact that divergent...
To improve crop model performance for regional yield estimates, a new four-dimensional variational algorithm (POD4DVar) merging the Monte Carlo and proper orthogonal decomposition techniques was introduced to develop data assimilation strategy using Crop Environment Resource Synthesis (CERES)-Wheat model. Two winter wheat estimation procedures were conducted on field plot scale test feasibility potential of POD4DVar-based strategy. Winter forecasts plots showed coefficient determination (R2)...
As an essential communication way for deaf-mutes, sign languages are expressed by human actions. To distinguish actions language understanding, the skeleton which contains position information of pose can provide important cue, since different usually correspond to poses/skeletons. However, has not been fully studied Sign Language Translation (SLT), especially end-to-end SLT. Therefore, in this paper, we propose a novel Skeleton-Aware neural Network (SANet) video-based Specifically, achieve...