A5086590282- Radiomics and Machine Learning in Medical Imaging
- Medical Imaging Techniques and Applications
- Lung Cancer Diagnosis and Treatment
- Machine Learning and ELM
- Radiopharmaceutical Chemistry and Applications
- Cancer Immunotherapy and Biomarkers
- Ultrasound and Hyperthermia Applications
- Cardiac Imaging and Diagnostics
- Advanced Radiotherapy Techniques
- Robotics and Sensor-Based Localization
- Dupuytren's Contracture and Treatments
- Lung Cancer Research Studies
- Autoimmune Bullous Skin Diseases
- Indoor and Outdoor Localization Technologies
- Cancer therapeutics and mechanisms
- Metastasis and carcinoma case studies
- Cancer Genomics and Diagnostics
- Recommender Systems and Techniques
- Acne and Rosacea Treatments and Effects
- Cancer Cells and Metastasis
- Sparse and Compressive Sensing Techniques
- Image Processing Techniques and Applications
- Underwater Vehicles and Communication Systems
- Domain Adaptation and Few-Shot Learning
- Data Stream Mining Techniques
Northeastern University
2023-2025
Beijing University of Chinese Medicine
2025
Shandong University
2014-2023
Shandong Provincial Hospital
2014-2023
Shandong First Medical University
2014-2022
Shandong Tumor Hospital
2014-2022
Academy of Medical Sciences
2021
Shandong Provincial Institute of Dermatology and Venereology
2019
Obstetrics and Gynecology Hospital of Fudan University
2014
Chinese Academy of Medical Sciences & Peking Union Medical College
2014
The goal of continual learning (CL) is to learn from a series continuously arriving new tasks without forgetting previously learned old tasks. To avoid catastrophic tasks, orthogonal gradient projection (OGP) based CL methods constrain the gradients be space spanned by This strict constraint will limit ability resulting in lower performance on In this paper, we first establish unified framework for OGP-based methods. We then revisit perspective loss landscape, where find that when relaxing...
Graph Contrastive Learning (GCL) leverages data augmentation techniques to produce contrasting views, enhancing the accuracy of recommendation systems through learning consistency between contrastive views. However, existing methods, such as directly perturbing interaction graph (e.g., node/edge dropout), may interfere with original connections and generate poor resulting in sub-optimal performance. In this paper, we define views that share only a small amount information due noisy (i.e.,...
Funding sources: none. Conflicts of interest: none to declare. Dear Editor, A 5‐year‐old boy came dermatology with tense blisters on his face, trunk and extremities 2 months' duration. All fingernail plates had separated from the nail matrices (a). diagnosis linear IgA bullous dermatosis was made based pathological immunopathological findings (c). The lesions were cleared a 4‐month course dapsone methylprednisolone (b). To our knowledge, this is first reported case coexisting onychomadesis....
Continual learning (CL) aims to learn new tasks without forgetting previous tasks. However, existing CL methods require a large amount of raw data, which is often unavailable due copyright considerations and privacy risks. Instead, stakeholders usually release pre-trained machine models as service (MLaaS), users can access via APIs. This paper considers two practical-yet-novel settings: data-efficient (DECL-APIs) data-free (DFCL-APIs), achieve from stream APIs with partial or no data....
Introduction: Follicular lymphoma (FL) is the most frequently occurring indolent lymphoma, and progression/relapse of disease within 24 months (POD24) now established as a predictor for poor survival, being leading cause death. Machine learning (ML) offers novel technology has several advantages compared with conventional models. We aimed to establish comprehensive model which could predict early POD (E-POD) FL patients. Methods: Data from retrospective cohort 139 patients (training cohort)...
Introduction: The efficacy and safety of chidamide, a novel subtype-selective histone deacetylase (HDAC) inhibitor, have been demonstrated in pivotal phase 2 clinical trial (CHIPEL study) were recently approved by China Food Drug Administration (CFDA) for the treatment patients with relapsed or refractory peripheral T cell lymphoma (PTCL). purpose this post-marketing surveillance study was to further elucidate real-world utilization chidamide PTCL. Methods: Patients confirmed PTCL who...
To detect the candidate genes for preeclampsia (PE).The gene expression profiles in preeclamptic and normal placental tissues were analyzed using cDNA microarray approach altered of important further confirmed by real-time RT-PCR (reverse transcription polymerase chain reaction) technique. Total RNA was extracted from three cases with severe PE pregnancy. After scanning, differentially expressed detected software.In two experiments (the fluorescent labels exchanged), a total 111 detected. In...
Background: Several studies have documented that FB23-2, as the selective inhibitor of RNA N6-methyladenosine (m6A) demethylase FTO, exerts crucial role in dampening tumorigenesis. Indeed, FTO deactivation was proved to render resistant leukemia cells sensitive targeted agents, which suggests a potential combination therapy strategy FB23-2. Aims: The present study aimed identify synergistic utility FB23-2 with ibrutinib CLL. Methods: A disseminated xenograft murine model established explore...
Background: Acetyl-CoA acetyltransferase 1 (ACAT1) is an enzyme to catalyze reversible formation of acetoacetyl-CoA. Yet, the roles ACAT1 in malignant progression have been incompletely understood so far. Hence, aim this study was investigate function and significance tumors. Methods: In present study, clinically annotated tumor patients from multiple cohorts were enrolled with informed consents. Gene expression analysis completed through TIMER2 web GEPIA2 web, based on combination HPA,...
Introduction: Increasing studies have identified that non-coding RNAs play critical roles in the initiation and progression of tumors via participating competitive endogenous RNA (ceRNA) networks. However, functions ceRNA network chronic lymphocytic leukemia (CLL) remain poorly understood. This study aims to explore molecular mechanism CLL provide potential prognostic markers therapeutic targets through integrated analysis CLL. Methods: The expression profile RNAS patients, cell lines (MEC-1...