检索结果-内蒙古大学图书馆

12th International Conference on Learning Representations, ICLR 2024

作者： Blumberg, Stefano B. Slator, Paddy J. Alexander, Daniel C. Centre for Artificial Intelligence Department of Computer Science University College London United Kingdom Centre for Medical Image Computing Department of Computer Science University College London United Kingdom Cardiff University Brain Research Imaging Centre School of Computer Science Cardiff University United Kingdom

This paper presents a data-driven, task-specific paradigm for experimental design, to shorten acquisition time, reduce costs, and accelerate the deployment of imaging devices. Current approaches in experimental design focus on model-parameter estimation and require specification of a particular model, whereas in imaging, other tasks may drive the design. Furthermore, such approaches often lead to intractable optimization problems in real-world imaging applications. Here we present a new paradigm for experimental design that simultaneously optimizes the design (set of image channels) and trains a machine-learning model to execute a user-specified image-analysis task. The approach obtains data densely-sampled over the measurement space (many image channels) for a small number of acquisitions, then identifies a subset of channels of prespecified size that best supports the task. We propose a method: TADRED for TAsk-DRiven Experimental Design in imaging, to identify the most informative channel-subset whilst simultaneously training a network to execute the task given the subset. Experiments demonstrate the potential of TADRED in diverse imaging applications: several clinically-relevant tasks in magnetic resonance imaging;and remote sensing and physiological applications of hyperspectral imaging. Results show substantial improvement over classical experimental design, two recent application-specific methods within the new paradigm, and state-of-the-art approaches in supervised feature selection. We anticipate further applications of our approach. Code is available: Code Link. © 2024 12th International Conference on Learning Representations, ICLR 2024. All rights reserved.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

A Two-Stage Self-Supervised Learning Framework for Automated Whole Heart Segmentation in CT and MRI: Addressing Challenges in Cardiac Imaging 1st

A Two-Stage Self-Supervised Learning Framework for Automated...

引用

1st MICCAI Challenge Comprehensive Analysis and computing of Real-World medical images, CARE 2024 Held in Conjunction with 27th International Conference on medical image computing and computer-Assisted Intervention, MICCAI 2024

作者： Qayyum, Abdul Mazher, Moona Niederer, Steven A. Faculty of Medicine National Heart and Lung Institute Imperial College London London United Kingdom Centre for Medical Image Computing Department of Computer Science University College London London United Kingdom

ISBN: (纸本)9783031870088

Cardiovascular diseases (CVDs) are the leading cause of death worldwide, highlighting the need for precise diagnostic and therapeutic strategies. Whole heart segmentation (WHS) from medical images is vital for understanding cardiac anatomy, assessing disease progression, and guiding treatments. However, WHS poses significant challenges due to the dynamic shape changes of the heart during the cardiac cycle, motion artifacts, low contrast-to-noise ratio, and variability from multi-center data. Additionally, CT and MRI modalities each present unique difficulties: CT provides high spatial resolution but poor soft-tissue contrast, while MRI offers excellent soft-tissue contrast but is prone to artifacts. To address these challenges, we propose a two-stage method. In the first stage, the DINOv2 framework leverages a 3D Vision-LSTM (xLSTM) model for self-supervised learning on the limited, unlabeled WHS++ dataset to overcome the scarcity of annotated data. In the second stage, we fine-tune the model for segmentation by freezing the encoder and optimizing the decoder. Our approach is rigorously evaluated on the WHS++ dataset for both CT and MRI modalities, demonstrating superior performance compared to state-of-the-art deep learning models in accurately segmenting cardiac structures. Our proposed solution achieved an average Dice Coefficient of 0.9414 using CT and 0.8891 using MRI dataset. This work advances the field of cardiac imaging by providing a robust, generalizable solution for automated whole heart segmentation. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： computerized tomography

来源：评论

学校读者我要写书评

暂无评论

Disentangled Diffusion Autoencoder for Harmonization of Multi-site Neuroimaging Data

arXiv

引用

arXiv 2024年

作者： Ijishakin, Ayodeji Aguila, Ana Lawry Levitis, Elizabeth Abdulaal, Ahmed Altmann, Andre Cole, James Centre for Medical Image Computing Department of Computer Science University College London United Kingdom

Combining neuroimaging datasets from multiple sites and scanners can help increase statistical power and thus provide greater insight into subtle neuroanatomical effects. However, site-specific effects pose a challenge by potentially obscuring the biological signal and introducing unwanted variance. Existing harmonization techniques, which use statistical models to remove such effects, have been shown to incompletely remove site effects while also failing to preserve biological variability. More recently, generative models using GANs or autoencoder-based approaches, have been proposed for site adjustment. However, such methods are known for instability during training or blurry image generation. In recent years, diffusion models have become increasingly popular for their ability to generate high-quality synthetic images. In this work, we introduce the disentangled diffusion autoencoder (DDAE), a novel diffusion model designed for controlling specific aspects of an image. We apply the DDAE to the task of harmonizing MR images by generating high-quality site-adjusted images that preserve biological variability. We use data from 7 different sites and demonstrate the DDAE’s superiority in generating high-resolution, harmonized 2D MR images over previous approaches. As far as we are aware, this work marks the first diffusion-based model for site adjustment of neuroimaging data. © 2024, CC BY-NC-SA.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

Efficient Deep Learning Models for Ultra-widefield Fundus Imaging for Diabetic Retinopathy 1st

Efficient Deep Learning Models for Ultra-widefield Fundus Im...

引用

1st MICCAI Challenge on Ultra-Widefield Fundus Imaging for Diabetic Retinopathy, UWF4DR 2024, Held in Conjunction with 27th International Conference on medical image computing and computer Assisted Intervention, MICCAI 2024

作者： Qayyum, Abdul Mazher, Moona Niederer, Steven A. National Heart and Lung Institute Faculty of Medicine Imperial College London London United Kingdom Centre for Medical Image Computing Department of Computer Science University College London London United Kingdom

ISBN: (纸本)9783031893872

Diabetic retinopathy (DR) is a leading cause of preventable blindness among working-age adults, with global cases expected to rise from 103 million in 2020 to 161 million by 2045. Early detection and treatment are essential for preventing vision loss. While color fundus photography (CFP) is the gold standard for DR diagnosis, ultra-widefield (UWF) fundus imaging, which provides up to a 200-degree view of the retina, is gaining traction as an alternative. However, classifying UWF images remains labor-intensive, and research on automated analysis is limited. In this study, we trained several deep-learning models for DR detection and assessment using UWF imaging. For Task 1, image Quality Assessment, we utilized the DINOv2 models due to their ability to efficiently manage complex features while maintaining computational efficiency. For Task 2, Identification of Referable DR, we employed MobileNetV3, which stood out for its lightweight architecture and advanced features like Squeeze-and-Excitation modules and H-Swish activation. MobileNetV3 achieved high accuracy with minimal computational requirements, making it well-suited for clinical deployment in resource-limited environments. To further enhance the model performance, we applied test-time augmentation (TTA) during inference, improving robustness and accuracy in challenging cases. This approach offers a powerful and efficient solution for DR classification and detection in UWF fundus images. Our proposed model ranked second in identifying referable DR on the final test dataset. Souce code is available at https://***/RespectKnowledge/UWF4DR_Challenge_2nd_place. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Alternative Learning Paradigms for image Quality Transfer

arXiv

引用

arXiv 2024年

作者： Eldaly, Ahmed Karam Figini, Matteo Alexander, Daniel C. Centre for Medical Image Computing Department of Computer Science University College London United Kingdom Department of Computer Science University of Exeter United Kingdom

image Quality Transfer (IQT) aims to enhance the contrast and resolution of low-quality medical images, e.g. obtained from low-power devices, with rich information learned from higher quality images. In contrast to existing IQT methods in the literature which adopt supervised learning frameworks, in this work, we propose two novel formulations of the IQT problem. The first approach uses an unsupervised learning framework, whereas the second is a combination of both supervised and unsupervised learning. The unsupervised learning approach considers a sparse representation (SRep) and dictionary learning model, which we call IQT-SRep, whereas the combination of supervised and unsupervised learning approach is based on deep dictionary learning (DDL), which we call IQT-DDL. The IQT-SRep approach trains two dictionaries using a sparse representation model using pairs of low- and high-quality volumes. Subsequently, the sparse representation of a low-quality block, in terms of the low-quality dictionary, can be directly used to recover the corresponding high-quality block using the high-quality dictionary. On the other hand, the IQT-DDL approach explicitly learns a high-resolution dictionary to upscale the input volume, while the entire network, including high dictionary generator, is simultaneously optimised to take full advantage of deep learning methods. The two models are evaluated using a low-field magnetic resonance imaging (MRI) application aiming to recover high-quality images akin to those obtained from high-field scanners. Experiments comparing the proposed approaches against state-of-the-art supervised deep learning IQT method (IQT-DL) identify that the two novel formulations of the IQT problem can avoid bias associated with supervised methods when tested using out-of-distribution data that differs from the distribution of the data the model was trained on. This highlights the potential benefit of these novel paradigms for IQT. © 2024, CC BY.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

LEARNING TO DOWNSAMPLE FOR SEGMENTATION OF ULTRA-HIGH RESOLUTION imageS 10

LEARNING TO DOWNSAMPLE FOR SEGMENTATION OF ULTRA-HIGH RESOLU...

引用

10th International Conference on Learning Representations, ICLR 2022

作者： Jin, Chen Tanno, Ryutaro Mertzanidou, Thomy Panagiotaki, Eleftheria Alexander, Daniel C. Centre for Medical Image Computing Department of Computer Science University College London United Kingdom Healthcare Intelligence Microsoft Research Cambridge United Kingdom

Many computer vision systems require low-cost segmentation algorithms based on deep learning, either because of the enormous size of input images or limited computational budget. Common solutions uniformly downsample the input images to meet memory constraints, assuming all pixels are equally informative. In this work, we demonstrate that this assumption can harm the segmentation performance because the segmentation difficulty varies spatially (see Figure 1 "Uniform"). We combat this problem by introducing a learnable downsampling module, which can be optimised together with the given segmentation model in an end-to-end fashion. We formulate the problem of training such downsampling module as optimisation of sampling density distributions over the input images given their low-resolution views. To defend against degenerate solutions (e.g. over-sampling trivial regions like the backgrounds), we propose a regularisation term that encourages the sampling locations to concentrate around the object boundaries. We find the downsampling module learns to sample more densely at difficult locations, thereby improving the segmentation performance (see Figure 1 "Ours"). Our experiments on benchmarks of high-resolution street view, aerial and medical images demonstrate substantial improvements in terms of efficiency-and-accuracy trade-off compared to both uniform downsampling and two recent advanced downsampling techniques. © 2022 ICLR 2022 - 10th International Conference on Learning Representationss. All rights reserved.

关键词： Antennas

来源：评论

学校读者我要写书评

暂无评论

Interpretable Alzheimer's Disease Classification Via a Contrastive Diffusion Autoencoder

arXiv

引用

arXiv 2023年

作者： Ijishakin, Ayodeji Abdulaal, Ahmed Hadjivasiliou, Adamos Martin, Sophie Cole, James Centre for Medical Image Computing Department of Computer Science University College London United Kingdom

In visual object classification, humans often justify their choices by comparing objects to prototypical examples within that class. We may therefore increase the interpretability of deep learning models by imbuing them with a similar style of reasoning. In this work, we apply this principle by classifying Alzheimer's Disease based on the similarity of images to training examples within the latent space. We use a contrastive loss combined with a diffusion autoencoder backbone, to produce a semantically meaningful latent space, such that neighbouring latents have similar image-level features. We achieve a classification accuracy comparable to black box approaches on a dataset of 2D MRI images, whilst producing human interpretable model explanations. Therefore, this work stands as a contribution to the pertinent development of accurate and interpretable deep learning within medical imaging. Code available at: Contrast-DiffAE. © 2023, CC BY-NC-ND.

关键词： Neurodegenerative diseases

来源：评论

学校读者我要写书评

暂无评论

POLAFFINI: Efficient feature-based polyaffine initialization for improved non-linear image registration

arXiv

引用

arXiv 2024年

作者： Legouhy, Antoine Callaghan, Ross Azadbakht, Hojjat Zhang, Hui Centre for Medical Image Computing Department of Computer Science University College London London United Kingdom AINOSTICS ltd. Manchester United Kingdom

This paper presents an efficient feature-based approach to initialize non-linear image registration. Today, nonlinear image registration is dominated by methods relying on intensity-based similarity measures. A good estimate of the initial transformation is essential, both for traditional iterative algorithms and for recent one-shot deep learning (DL)-based alternatives. The established approach to estimate this starting point is to perform affine registration, but this may be insufficient due to its parsimonious, global, and non-bending nature. We propose an improved initialization method that takes advantage of recent advances in DL-based segmentation techniques able to instantly estimate fine-grained regional delineations with state-of-the-art accuracies. Those segmentations are used to produce local, anatomically grounded, featurebased affine matchings using iteration-free closed-form expressions. Estimated local affine transformations are then fused, with the log-Euclidean polyaffine framework, into an overall dense diffeomorphic transformation. We show that, compared to its affine counterpart, the proposed initialization leads to significantly better alignment for both traditional and DL-based non-linear registration algorithms. The proposed approach is also more robust and significantly faster than commonly used affine registration algorithms such as FSL FLIRT. © 2024, CC BY.

关键词： image registration

来源：评论

学校读者我要写书评

暂无评论

Feature Attention as a Control Mechanism for the Balance of Speed and Accuracy in Visual Search

引用

Computational Brain and Behavior 2023年第3期6卷 503-512页

作者： Griffith, Thom Townend, Florence J. Baker, Sophie-Anne Lepora, Nathan F. Department of Engineering Maths University of Bristol University Walk Bristol BS8 1TW United Kingdom Centre for Medical Image Computing Department of Computer Science University College London Gower Street London WC1E 6BT United Kingdom

Finding an object amongst a cluttered visual scene is an everyday task for humans but presents a fundamental challenge to computational models performing this feat. Previous attempts to model efficient visual search have focused on locating targets as swiftly as possible, but so far have not considered balancing the costs of lengthy searches against the costs of making errors. Here, we propose a neuro-inspired model of visual search that offers an attention-based control mechanism for this speed-accuracy trade-off. The model combines a goal-based fixation policy, which captures human-like behaviour on a simple visual search task, with a deep neural network that carries out the target detection step. The neural network is patched with a target-based feature attention model previously applied to standalone classification tasks. In contrast to image classification, visual search introduces a time component, which places an additional demand on the model to minimise the time cost of the search whilst also maintaining acceptable accuracy. The proposed model balances these two costs by modulating the attentional strength given to characteristic features of the target class, thereby minimising an associated cost function. The model offers a method for optimising the costs of visual search and demonstrates the value of a decision theoretic approach to modelling more complex visual tasks involving attention. © 2023, The Author(s).

关键词： CNN Feature attention Speed-accuracy trade-off Visual search

来源：评论

学校读者我要写书评

暂无评论

SCREENER: A general framework for task-specific experiment design in quantitative MRI

arXiv

引用

arXiv 2024年

作者： Zheng, Tianshu Wang, Zican Bray, Timothy Alexander, Daniel C. Wu, Dan Zhang, Hui College of Biomedical Engineering & Instrument Science Zhejiang University Zhejiang Hangzhou China Centre for Medical Image Computing Department of Computer Science University College London London United Kingdom

Quantitative magnetic resonance imaging (qMRI) is increasingly investigated for use in a variety of clinical tasks from diagnosis, through staging, to treatment monitoring. However, experiment design in qMRI, the identification of the optimal acquisition protocols, has been focused on obtaining the most precise parameter estimations, with no regard for the specific requirements of downstream tasks. Here we propose SCREENER: A general framework for task-specific experiment design in quantitative MRI. SCREENER incorporates a task-specific objective and seeks the optimal protocol with a deep-reinforcement-learning (DRL) based optimization strategy. To illustrate this framework, we employ a task of classifying the inflammation status of bone marrow using diffusion MRI data with intravoxel incoherent motion (IVIM) modelling. Results demonstrate SCREENER outperforms previous ad hoc and optimized protocols under clinical signal-to-noise ratio (SNR) conditions, achieving significant improvement, both in binary classification tasks, e.g. from 67% to 89%, and in a multi-class classification task, from 46% to 59%. Additionally, we show this improvement is robust to the SNR. Lastly, we demonstrate the advantage of DRL-based optimization strategy, enabling zero-shot discovery of near-optimal protocols for a range of SNRs not used in training. In conclusion, SCREENER has the potential to enable wider uptake of qMRI in the clinic. Copyright © 2024, The Authors. All rights reserved.

关键词： Signal to noise ratio

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：