A5074960174- Radiomics and Machine Learning in Medical Imaging
- Advanced Image and Video Retrieval Techniques
- Medical Imaging Techniques and Applications
- Advanced Neural Network Applications
- Medical Image Segmentation Techniques
- MRI in cancer diagnosis
- Glioma Diagnosis and Treatment
- Remote-Sensing Image Classification
- Lung Cancer Diagnosis and Treatment
- Multimodal Machine Learning Applications
- AI in cancer detection
- Advanced X-ray and CT Imaging
- Hepatocellular Carcinoma Treatment and Prognosis
- Domain Adaptation and Few-Shot Learning
- Ovarian cancer diagnosis and treatment
- Optical measurement and interference techniques
- Robotics and Sensor-Based Localization
- Medical Imaging and Analysis
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Advanced X-ray Imaging Techniques
- Color perception and design
- Visual perception and processing mechanisms
- Brain Metastases and Treatment
- Generative Adversarial Networks and Image Synthesis
- Infrared Target Detection Methodologies
Suzhou Institute of Biomedical Engineering and Technology
2016-2025
Chinese Academy of Sciences
2016-2025
China Three Gorges University
2025
Zhejiang Provincial People's Hospital
2023-2025
Hangzhou Medical College
2023-2025
Changchun Institute of Optics, Fine Mechanics and Physics
2025
University of Chinese Academy of Sciences
2024-2025
Nantong University
2024-2025
Aerospace Information Research Institute
2020-2024
China Medical University
2024
Accurate prediction of treatment response to neoadjuvant chemotherapy (NACT) in individual patients with locally advanced gastric cancer (LAGC) is essential for personalized medicine. We aimed develop and validate a deep learning radiomics nomogram (DLRN) based on pretreatment contrast-enhanced computed tomography (CT) images clinical features predict the NACT LAGC.719 LAGC were retrospectively recruited from four Chinese hospitals between Dec 1st, 2014 Nov 30th, 2020. The training cohort...
Increasing demand for the knowledge about protein-protein interactions (PPIs) is promoting development of methods predicting protein interaction network. Although high-throughput technologies have generated considerable PPIs data various organisms, it has inevitable drawbacks such as high cost, time consumption, and inherently false positive rate. For this reason, computational are drawing more attention PPIs. In study, we report a method using information sequences. The main improvements...
Semantic segmentation of aerial images refers to assigning one land cover category each pixel. This is a challenging task due the great differences in appearances ground objects. Many attempts have been made during past decades. In recent years, convolutional neural networks (CNNs) introduced remote sensing field, and various solutions proposed realize dense semantic labeling with CNNs. this letter, we propose shuffling CNNs periodic manner. approach supplement current methods for images. We...
BackgroundAccurate lymph nodes (LNs) assessment is important for rectal cancer (RC) staging in multiparametric magnetic resonance imaging (mpMRI). However, it incredibly time-consumming to identify all the LNs scan region. This study aims develop and validate a deep-learning-based, fully-automated node detection segmentation (auto-LNDS) model based on mpMRI.MethodsIn total, 5789 annotated (diameter ≥ 3 mm) mpMRI from 293 patients with RC single center were enrolled. Fused T2-weighted images...
The balance between high accuracy and speed has always been a challenging task in semantic image segmentation. Compact segmentation networks are more widely used the case of limited resources, while their performances constrained. In this paper, motivated by residual learning global aggregation, we propose simple yet general effective knowledge distillation framework called double similarity (DSD) to improve classification all existing compact capturing pixel category dimensions,...
Abstract Background The tumor immune microenvironment can influence the prognosis and treatment response to immunotherapy. We aimed develop a non-invasive radiomic signature in high-grade glioma (HGG) predict absolute density of tumor-associated macrophages (TAMs), preponderant cells HGG. also evaluate association between signature, phenotype as well Methods In this retrospective setting, total 379 patients with HGG from three independent cohorts were included construct model named Radiomics...
Synthetic aperture radar (SAR) ship detection plays an important role in the field of maritime security. However, certain unique imaging properties make it challenging to extract shape features ships, such as speckle noise and strong scattering interference from irrelevant objects. These factors result inaccurate localization obvious false alarms under complex large-scale inshore scenes. To address this issue, we propose region topology network (SRT-Net), which can dynamically capture...
This study presents a multimodal machine learning model to predict ICD-10 diagnostic codes. We developed separate models that can handle data from different modalities, including unstructured text, semi-structured text and structured tabular data. further employed an ensemble method integrate all modality-specific generate Key evidence was also extracted make our prediction more convincing explainable. used the Medical Information Mart for Intensive Care III (MIMIC -III) dataset validate...
Objectives: To establish a radiomic algorithm based on grayscale ultrasound images and to make preoperative predictions of microvascular invasion (MVI) in hepatocellular carcinoma (HCC) patients. Methods: In this retrospective study, 322 cases histopathologically confirmed HCC lesions were included. The classifications performed two stages: (1) classifier #1, MVI-negative MVI-positive cases; (2) #2, further classified as M1 or M2 cases. gross-tumoral region (GTR) peri-tumoral (PTR)...
The encoder–decoder framework is prevalent in existing remote-sensing image captioning (RSIC) models. appearance of attention mechanisms brings significant results. However, current attention-based caption models only build up the relationships between local features without introducing global visual feature and removing redundant components. It will cause to generate descriptive sentences that are weakly related scene images. To solve problems, this article proposed a feature-guided (GVFGA)...
Preoperative differentiation of borderline from malignant epithelial ovarian tumors (BEOT MEOT) can impact surgical management. MRI has improved this assessment but subjective interpretation by radiologists may lead to inconsistent results.To develop and validate an objective MRI-based machine-learning (ML) model for differentiating BEOT MEOT, compare the performance against radiologists' interpretation.Retrospective study eight clinical centers.In all, 501 women with...
Background In unresectable hepatocellular carcinoma (HCC), methods to predict patients at increased risk of progression are required. Purpose To investigate the feasibility radiomics model in predicting early HCC after transcatheter arterial chemoembolization (TACE) therapy using preoperative multiparametric magnetic resonance imaging (MP‐MRI). Study Type Retrospective. Population A total 84 with BCLC B stage from one medical center. According modified response evaluation criteria solid...
Abstract Purpose To investigate the performance of magnetic resonance imaging (MRI)-based radiomics models for benign and malignant prostate lesion discrimination extracapsular extension (ECE) positive surgical margins (PSM) prediction. Methods materials In total, 459 patients who underwent multiparametric MRI (mpMRI) before biopsy were included. Radiomic features extracted from both T2-weighted (T2WI) apparent diffusion coefficient (ADC). Patients divided into different training sets...
Preoperative prediction of extracapsular extension (ECE) prostate cancer (PCa) is important to guide clinical decision-making and improve patient prognosis.To investigate the value multiparametric magnetic resonance imaging (mpMRI)-based peritumoral radiomics for preoperative presence ECE.Retrospective.Two hundred eighty-four patients with PCa from two centers (center 1: 226 patients; center 2: 58 patients). Cases 1 were randomly divided into training (158 patients) internal validation (68...
Much of the recent work in remote sensing image captioning is influenced by natural captioning. These methods tend to fix defects model architecture improve previous work, but pay little attention differences between images and images. By considering these differences, we propose a multiscale multiinteraction model. As Fig. 1(a), targets have wide range scales; while are generally taken close-up, resulting similar scale for foreground targets. Due difference shooting methods, pretrained on...
Purpose To develop a Weakly supervISed model DevelOpment fraMework (WISDOM) to construct lymph node (LN) diagnosis for patients with rectal cancer (RC) that uses preoperative MRI data coupled postoperative patient-level pathologic information. Materials and Methods In this retrospective study, the WISDOM was built using (T2-weighted diffusion-weighted imaging) information (the number of postoperatively confirmed metastatic LNs resected LNs) based on RC between January 2016 November 2017. The...
Protein-protein interactions are the basis of biological functions, and studying these on a molecular level is crucial importance for understanding functionality living cell. During past decade, biosensors have emerged as an important tool high-throughput identification proteins their interactions. However, experimental methods identifying PPIs both time-consuming expensive. On other hand, PPI data often associated with high false-positive false-negative rates. Targeting at problems, we...
Our objective was to identify prognostic imaging biomarkers for hepatocellular carcinoma in contrast-enhanced computed tomography (CECT) with biological interpretations by associating features and gene modules.