检索结果-内蒙古大学图书馆

A Multi-Stage Framework for the 2022 Multi-Structure Segmentation for Renal Cancer Treatment

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Liu, Yusheng Zhao, Zhongchen Wang, Lisheng Institute of Image Processing and Pattern Recognition Department of Automation Shanghai Jiao Tong University Shanghai200240 China

Three-dimensional (3D) kidney parsing on computed tomography angiography (CTA) images is of great clinical significance. Automatic segmentation of kidney, renal tumor, renal vein and renal artery benefits a lot on surgery-based renal cancer treatment. In this paper, we propose a new nnhra-unet network, and use a multi-stage framework which is based on it to segment the multi-structure of kidney and participate in the KiPA2022 challenge. Copyright © 2022, The Authors. All rights reserved.

关键词： Diseases

Decentralized Kernel Ridge Regression Based on Data-Dependent Random Feature

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Yang, Ruikai He, Fan He, Mingzhen Yang, Jie Huang, Xiaolin The Institute of Image Processing and Pattern Recognition The MOE Key Laboratory of System Control and Information Processing Shanghai Jiao Tong University Shanghai200240 China The STADIUS Center for Dynamical Systems Signal Processing and Data Analytics KU Leuven LeuvenB-3001 Belgium

Random feature (RF) has been widely used for node consistency in decentralized kernel ridge regression (KRR). Currently, the consistency is guaranteed by imposing constraints on coefficients of features, necessitating that the random features on different nodes are identical. However, in many applications, data on different nodes varies significantly on the number or distribution, which calls for adaptive and data-dependent methods that generate different RFs. To tackle the essential difficulty, we propose a new decentralized KRR algorithm that pursues consensus on decision functions, which allows great flexibility and well adapts data on nodes. The convergence is rigorously given and the effectiveness is numerically verified: by capturing the characteristics of the data on each node, while maintaining the same communication costs as other methods, we achieved an average regression accuracy improvement of 25.5% across six real-world data sets. Copyright © 2024, The Authors. All rights reserved.

关键词： Consensus algorithm

MDFlow: Unsupervised Optical Flow Learning by Reliable Mutual Knowledge Distillation

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Kong, Lingtong Yang, Jie The Institute of Image Processing and Pattern Recognition Department of Automation Shanghai Jiao Tong University Shanghai200240 China

Recent works have shown that optical flow can be learned by deep networks from unlabelled image pairs based on brightness constancy assumption and smoothness prior. Current approaches additionally impose an augmentation regularization term for continual self-supervision, which has been proved to be effective on difficult matching regions. However, this method also amplify the inevitable mismatch in unsupervised setting, blocking the learning process towards optimal solution. To break the dilemma, we propose a novel mutual distillation framework to transfer reliable knowledge back and forth between the teacher and student networks for alternate improvement. Concretely, taking estimation of off-the-shelf unsupervised approach as pseudo labels, our insight locates at defining a confidence selection mechanism to extract relative good matches, and then add diverse data augmentation for distilling adequate and reliable knowledge from teacher to student. Thanks to the decouple nature of our method, we can choose a stronger student architecture for sufficient learning. Finally, better student prediction is adopted to transfer knowledge back to the efficient teacher without additional costs in real deployment. Rather than formulating it as a supervised task, we find that introducing an extra unsupervised term for multi-target learning achieves best final results. Extensive experiments show that our approach, termed MDFlow, achieves state-of-the-art real-time accuracy and generalization ability on challenging benchmarks. Code is available at https://***/ltkong218/MDFlow. Copyright © 2022, The Authors. All rights reserved.

关键词： Distillation

Detection of Venous Thromboembolism Using Recurrent Neural Networks with Time-Series Data 24

学校读者我要写书评

暂无评论

Detection of Venous Thromboembolism Using Recurrent Neural N...

3rd Asia Conference on Algorithms, Computing and Machine Learning, CACML 2024

作者： Xu, Can Huang, Yaqin Xiang, Xinni Lei, Haike Yang, Jie Shanghai Jiao Tong University Shanghai China Chongqing University Cancer Hospital Chongqing China West China Hospital Sichuan University Chengdu China Shanghai Jiao Tong University Institute of Image Processing and Pattern Recognition China Chongqing Key Laboratory of Translational Research for Cancer Metastasis and Individualized Treatment China West China School of Medicine China Chongqing Cancer Multi-omics Big Data Application Engineering Research Center China

ISBN: (纸本)9798400716416

Machine Learning (ML) has been widely applied to medical science for decades. As common knowledge, the progress of many diseases is often chronic and dynamic. Longitudinal data, or time-series data, has better descriptions of certain diseases, as it naturally captures the dynamic over time. Venous thromboembolism (VTE), as a regular complication of cancer tumour treatments, has been emphasized in many studies. Recent research has shown that ML algorithms incredibly contribute to VTE risk estimation and prediction. However, existing methods rarely consider the dynamic changes during the development of VTE, and often make decisions based on time-point data, such as the latest blood test results of patients. In this paper, we propose a VTE detection model that takes advantage of capturing dynamic information using time-series data. The model consists of a Recurrent Neural Network (RNN) based framework with time embedding and Attention mechanism to encode longitudinal features of patients. We test our proposed model on a real-world dataset that contains multiple times of blood test results of patients together with other clinical features. The network is compared with several commonly used ML models in previous VTE studies. The results show that our model outperforms all baseline models, and achieved a state-of-art performance with 80.0% accuracy and 0.881 AUC. © 2024 ACM.

关键词： Risk assessment

Distilling Object Detectors with Global Knowledge

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Tang, Sanli Zhang, Zhongyu Cheng, Zhanzhan Lu, Jing Xu, Yunlu Niu, Yi He, Fan Hikvision Research Institute Hangzhou China Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Shanghai China

Knowledge distillation learns a lightweight student model that mimics a cumbersome teacher. Existing methods regard the knowledge as the feature of each instance or their relations, which is the instance-level knowledge only from the teacher model, i.e., the local knowledge. However, the empirical studies show that the local knowledge is much noisy in object detection tasks, especially on the blurred, occluded, or small instances. Thus, a more intrinsic approach is to measure the representations of instances w.r.t. a group of common basis vectors in the two feature spaces of the teacher and the student detectors, i.e., global knowledge. Then, the distilling algorithm can be applied as space alignment. To this end, a novel prototype generation module (PGM) is proposed to find the common basis vectors, dubbed prototypes, in the two feature spaces. Then, a robust distilling module (RDM) is applied to construct the global knowledge based on the prototypes and filtrate noisy local knowledge by measuring the discrepancy of the representations in two feature spaces. Experiments with Faster-RCNN and RetinaNet on PASCAL and COCO datasets show that our method achieves the best performance for distilling object detectors with various backbones, which even surpasses the performance of the teacher model. We also show that the existing methods can be easily combined with global knowledge and obtain further improvement. © 2022, CC BY-NC-ND.

关键词： Distillation

Iteratively Refine the Segmentation of Head and Neck Tumor in FDG-PET and CT images 1

学校读者我要写书评

暂无评论

1st 3D Head and Neck Tumor Segmentation in PET/CT Challenge, HECKTOR 2020, which was held in conjunction with 23rd International Conference on Medical image Computing and Computer-Assisted Intervention, MICCAI 2020

作者： Chen, Huai Chen, Haibin Wang, Lisheng Department of Automation Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Shanghai200240 China Perception Vision Medical Technology Guangzhou China

ISBN: (数字)9783030671945

ISBN: (纸本)9783030671938

The automatic segmentation of head and neck (H&N) tumor from FDG-PET and CT images is urgently needed for radiomics. In this paper, we propose a framework to segment H&N tumor automatically by fusing information of PET and CT. In this framework, multiple 3D-Unets are trained one-by-one. The predictions and features of upstream models will be captured as additional information for the next one to further refine the segmentation. Experiments show that iterative refinements can improve the performance. We evaluated our framework on the dataset of HECKTOR2020 (The challenge of head and neck tumor segmentation) and won the 5th place with average DSC, precision and recall of 0.7241, 0.8479, 0.6701 respectively. © 2021, Springer Nature Switzerland AG.

关键词： Computerized tomography

Semantic Transformation-Based Data Augmentation for Few-Shot Learning

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Pan, Mei-Hong Xin, Hong-Yi Shen, Hong-Bin Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Key Laboratory of System Control and Information Processing Ministry of Education of China Shanghai200240 China Shanghai Jiaotong University Shanghai200240 China

Few-shot learning (FSL) as a data-scarce method, aims to recognize instances of unseen classes solely based on very few examples. However, the model can easily become overfitted due to the biased distribution formed with extremely limited training data. This paper presents a task specific data augmentation approach by transferring samples from base dataset to the novel tasks in an encoder-decoder paradigm, which guarantees generating semantically meaningful features. Specifically, the feature transfer process is carried out in semantic space. We further impose a compactness constraint to the generated features with the prototypes working as the reference points, which ensures the generated features distribute around the class centers. Moreover, we incorporate the cluster centers of the query set with the prototypes of the support set to reduce the bias of the class centers. With the supervision of the compactness loss, the model is encouraged to generate discriminative features with high inter-class dispersion and intra-class compactness. Extensive experiments show that our method outperforms the state-of-the-arts on four benchmarks, namely MiniimageNet, TieredimageNet, CUB and CIFAR-FS. © 2023, The Authors. All rights reserved.

关键词： Motion planning

3D Vessel Segmentation with Limited Guidance of 2D Structure-agnostic Vessel Annotations

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Chen, Huai Wang, Xiuying Wang, Lisheng Institute of Image Processing and Pattern Recognition Department of Automation Shanghai Jiao Tong University Shanghai200240 China The School of Computer Science The University of Sydney SydneyNSW2006 Australia

Delineating 3D blood vessels is essential for clinical diagnosis and treatment, however, is challenging due to complex structure variations and varied imaging conditions. Supervised deep learning has demonstrated its superior capacity in automatic 3D vessel segmentation. However, the reliance on expensive 3D manual annotations and limited capacity for annotation reuse hinder the clinical applications of supervised models. To avoid the repetitive and laborious annotating process and make full use of existing vascular annotations, this paper proposes a novel 3D shape-guided local discrimination (3D-SLD) model for 3D vascular segmentation under limited guidance from public 2D vessel annotations. The primary hypothesis is that 3D vessels are composed of semantically similar voxels and often exhibit tree-shaped morphology. Accordingly, the 3D region discrimination loss is firstly proposed to learn the discriminative representation measuring voxel-wise similarities and cluster semantically consistent voxels to form the candidate 3D vascular segmentation in unlabeled images;secondly, based on the similarity of the tree-shaped morphology between 2D and 3D vessels, the Crop- and-Overlap strategy is presented to generate reference masks from the existing 2D structure-agnostic vessel annotations, which are fit for varied vascular structures, and the adversarial loss is introduced to guide the 3D vessels with the tree-shaped morphology;and thirdly, the temporal consistency loss is proposed to foster the training stability and keep the model updated smoothly. To further enhance the model’s robustness and reliability, the orientation-invariant CNN module and Reliability-Refinement algorithm are presented. Experimental results from the public 3D cerebrovascular and 3D arterial tree datasets demonstrate that our model achieves comparable effectiveness against nine deep learning supervised models. Copyright © 2023, The Authors. All rights reserved.

关键词： Morphology

Pick the Best Pre-trained Model: Towards Transferability Estimation for Medical image Segmentation

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Yang, Yuncheng Wei, Meng He, Junjun Yang, Jie Ye, Jin Gu, Yun Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Shanghai China Institute of Medical Robotics Shanghai Jiao Tong University Shanghai China Shanghai AI Lab Shanghai China

Transfer learning is a critical technique in training deep neural networks for the challenging medical image segmentation task that requires enormous resources. With the abundance of medical image data, many research institutions release models trained on various datasets that can form a huge pool of candidate source models to choose from. Hence, it’s vital to estimate the source models’ transferability (i.e., the ability to generalize across different downstream tasks) for proper and efficient model reuse. To make up for its deficiency when applying transfer learning to medical image segmentation, in this paper, we therefore propose a new Transferability Estimation (TE) method. We first analyze the drawbacks of using the existing TE algorithms for medical image segmentation and then design a source-free TE framework that considers both class consistency and feature variety for better estimation. Extensive experiments show that our method surpasses all current algorithms for transferability estimation in medical image segmentation. Code is available at here. Copyright © 2023, The Authors. All rights reserved.

关键词： Deep neural networks