检索结果-内蒙古大学图书馆

A stain color normalization with robust dictionary learning for breast cancer histological images processing

BIOMEDICAL SIGNAL PROCESSING AND CONTROL 2023年第1期85卷

作者： Tosta, Thaina A. Azevedo Freitas, Andre Dias de Faria, Paulo Rogerio Neves, Leandro Alves Martins, Alessandro Santana do Nascimento, Marcelo Zanchetta Univ Fed Sao Paulo Inst Sci & Technol Ave Cesare Mansueto Giulio Lattes 1201 BR-12247014 Sao Jose Dos Campos SP Brazil Univ Fed Uberlandia Inst Biomed Sci Dept Histol & Morphol Ave Amazonas S-N BR-38405320 Uberlandia MG Brazil Sao Paulo State Univ Dept Comp Sci & Stat R Cristovao Colombo 2265 BR-15054000 Sao Jose Do Rio Preto SP Brazil Fed Inst Triangulo Mineiro R Belarmino Vilela Junqueira S-N BR-38305200 Ituiutaba MG Brazil Univ Fed Uberlandia Fac Comp Sci Ave Joao Naves de Avila 2121 BR-38400902 Uberlandia MG Brazil

Microscopic analyses of tissue samples are crucial for confirming the diagnosis of breast cancer. The digitization of these samples has led to the development of computational systems that can assist pathologists. However, these systems may face limitations owing to color variations in the images. Normalization studies have been widely conducted to address these issues, but there is still a need for new proposals that take into account the biological properties of dyes and tissues. This study presents a novel method for normalizing hematoxylin and eosin-stained histological images by estimating the color appearance matrices and density maps of the stain. The proposed method offers contributions in terms of pixel selection and weight definition to improve the color estimation of histological images. Besides, to the best of our knowledge, no previous studies have evaluated normalized images considering both handcrafted and learning features. Breast cancer images with significant color variations were used to evaluate this approach and the results demonstrated its effectiveness and efficiency. The average values of FSIM, NIQE, and QSSIM were up to 0.9866, 3.4298, and 0.9655, respectively. Compared with other normalization techniques, the proposed method showed an increase of up to 5.9261, with the largest difference observed in the amount of noise added, as indicated by the NIQE metric. To determine the impact of normalization on feature extraction, the evaluations included an analysis of both color and deep-learned features. These experiments showed that all evaluated methods harmed the separation of breast cancer samples by color features. In contrast, the deep-learned features resulted in less complex classification problems, especially with the proposed normalization. This technique also reached one of the lowest processing times, nearly 6 s with the largest image from the databases.

关键词： H&E histological images analysis Color normalization sparse non-negative matrix factorization Dictionary learning Features analysis

来源：评论

学校读者我要写书评

暂无评论

An inertial block majorization minimization framework for nonsmooth nonconvex optimization

The Journal of Machine Learning Research

引用

The Journal of Machine Learning Research 2023年第1期24卷 636-676页

作者： Le Thi Khanh Hien Duy Nhat Phan Nicolas Gillis University of Mons Belgium Department of Mathematics and Statistics University of Massachusetts Lowell Department of Mathematics and Operational Research University of Mons Belgium

In this paper, we introduce TITAN, a novel inerTIal block majorizaTion minimizAtioN framework for nonsmooth nonconvex optimization problems. To the best of our knowledge, TITAN is the first framework of block-coordinate update method that relies on the majorization-minimization framework while embedding inertial force to each step of the block updates. The inertial force is obtained via an extrapolation operator that subsumes heavy-ball and Nesterov-type accelerations for block proximal gradient methods as special cases. By choosing various surrogate functions, such as proximal, Lipschitz gradient, Bregman, quadratic, and composite surrogate functions, and by varying the extrapolation operator, TITAN produces a rich set of inertial block-coordinate update methods. We study sub-sequential convergence as well as global convergence for the generated sequence of TITAN. We illustrate the effectiveness of TITAN on two important machine learning problems, namely sparse non-negative matrix factorization and matrix completion.

关键词： inertial method block coordinate method majorization minimization surrogate functions sparse non-negative matrix factorization matrix completion

来源：评论

学校读者我要写书评

暂无评论

Combining a forward supervised filter learning with a sparse NMF for breast cancer histopathological image classification

引用

Intelligence-Based Medicine 2024年 10卷 100174-100174页

作者： Karuppasamy, ArunaDevi Abdesselam, Abdelhamid zidoum, Hamza Hedjam, Rachid Al-Bahri, Maiya Department of Computing Sciences Gulf College Oman Department of Computer Science Sultan Qaboos University Oman Department of Computer Science Bishop's University Sherbrooke Quebec Canada Department of Pathology College of Medicine and Health Sciences Sultan Qaboos University Oman

Histopathological images play a important role in clinical diagnosis, particularly in identifying and assessing the severity of abnormal conditions like benign lesions and malignant tumors. Traditional machine learning techniques for processing histopathology images involve the extraction of manual features from these images, which is typically done with the assistance of industry experts. Recent advancements in Deep Learning (DL), especially with Convolutional Neural Networks (CNN), have enabled the automatic extraction of multi-level abstract features directly from raw data. This capability significantly enhances the performance of complex computer vision tasks. Classic CNN models like AlexNet and VggNet employ back-propagation algorithms to learn filters in the training phase. However, these algorithms demand large labeled datasets, resulting in extensive computational processing. Additionally, they often face the vanishing gradient problem, which can negatively impact the quality of the learning process. Besides, in many domains, acquiring enough labeled images for conducting properly the training phase is a real challenge. To address these challenges, a feed-forward propagation approach was proposed using non-negative matrix factorization(NMF). The NMF technique factorizes the input data into two latent factors (non-negative matrices). It has been shown that by enforcing constraints such as sparsity on the latent factors, dominant features that are mostly correlated with tumors types can be extracted. In this work, a novel model combining sparse NMF and Support Vector Machine (SVM) was developed for classifying histopathological images. We have derived a mathematical model of a novel feed-forward filter learning approach that combines sparse NMF (SNMF) and Support Vector Machine technique (SVM). The model was used to design and implement a feed-forward CNN classifier to classify histopathology images. This model has been evaluated on the histopathology images f

关键词： Breast cancer Feed-forward convolutional networks Filter learning Histopathological images Image classification sparse non-negative matrix factorization Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Semi-supervised transient noise suppression using OMLSA and SNMF algorithms

引用

APPLIED ACOUSTICS 2020年 170卷 107533-107533页

作者： Ullah, Rizwan Islam, Md Shohidul Ye, Zhongfu Asif, Muhammad Univ Sci & Technol China Dept Elect Engn & Informat Sci Natl Engn Lab Speech & Language Informat Proc Hefei Anhui Peoples R China Islamic Univ Dept Comp Sci & Engn Kushtia 7003 Bangladesh City Univ Sci & Informat Technol Dept Elect Engn Peshawar Kpk Pakistan

Transient interferences such as keystrokes, mouse clicks and hammering pose a significant challenge in the single channel speech enhancement due to their abrupt and non-continuous nature. Traditional noise suppression algorithms and even many non-stationary noise reduction algorithms do not adequately suppress transient interference. Therefore, in this work, we propose a semi-supervised single channel transient noise suppression method to effectively suppress the transient interference without significant audible distortion. The proposed algorithm consists of training and testing stages. In the training stage, the proposed technique first uses the optimally modified-log spectral amplitude (OMLSA) estimator to estimate the transient noise from the noisy speech signal. After that, we eliminate the residual speech components from the estimated noise obtained from OMLSA based on the correlation coefficient, by taking correlation between the estimated noise with the available clean speech data from the dataset passed through the voice activity detector for silence zones removal. Afterwards, we use this noise for training the noise dictionary in sparse non-negative matrix factorization. Clean speech data is used for speech dictionary training. In the enhancement stage, the dictionaries are fixed and concatenated, to obtain the corresponding activation matrices. The clean speech dictionary and the corresponding weight matrix are used to reconstruct the estimated speech. The experimental results reveal that the proposed algorithm provided better performance compared to other existing algorithms in the speech quality evaluation metrics. (C) 2020 Elsevier Ltd. All rights reserved.

关键词： Transient Noise Noise Suppression Optimally Modified-Log Spectral Amplitude Estimation sparse non-negative matrix factorization Voice Activity Detection

来源：评论

学校读者我要写书评

暂无评论

Unsupervised method for normalization of hematoxylin-eosin stain in histological images

引用

COMPUTERIZED MEDICAL IMAGING AND GRAPHICS 2019年 77卷 101646-000页

作者： Azevedo Tosta, Thaina A. de Faria, Paulo Rogerio Silva Servato, Joao Paulo Neves, Leandro Alves Roberto, Guilherme Freire Martins, Alessandro Santana do Nascimento, Marcelo Zanchetta Fed Univ ABC Ctr Math Comp & Cognit Av Estados 5001 BR-09210580 Santo Andre SP Brazil Univ Fed Uberlandia Inst Biomed Sci Dept Histol & Morphol Av Amazonas S-N BR-38405320 Uberlandia MG Brazil Univ Uberaba Sch Dent Av Nene Sabino 1801 BR-38055500 Uberaba MG Brazil Sao Paulo State Univ Dept Comp Sci & Stat R Cristovao Colombo 2265 BR-15054000 Sao Jose Do Rio Preto SP Brazil Univ Fed Uberlandia Fac Comp Sci Av Joao Naves de Avila 2121 BR-38400902 Uberlandia MG Brazil Fed Inst Triangulo Mineiro R Belarmino Vilela Junqueira S-N BR-38305200 Ituiutaba MG Brazil

Histological images stained with hematoxylin-eosin are widely used by pathologists for cancer diagnosis. However, these images can have color variations that highly influence the histological image processing techniques. To deal with this potential limitation, normalization methods are useful for color correction. In this paper, a histological image color normalization is presented by considering the biological and hematoxylin-eosin properties. To this end, the stain representation of a reference image was applied in place of the original images representation, allowing the preservation of histological structures. This proposal was evaluated on histological images with great variations of contrast, and both visual and quantitative analyzes yielded promising results. (C) 2019 Elsevier Ltd. All rights reserved.

关键词： Histopathological image analysis Normalization Spectral matching sparse non-negative matrix factorization

来源：评论

学校读者我要写书评

暂无评论

Genome-Wide SNPs Identify Limits to Connectivity in the Extreme Freshwater Disperser, Spangled Perch Leiopotherapon unicolor (Terapontidae)

引用

JOURNAL OF HEREDITY 2018年第3期109卷 320-325页

作者： Schmidt, Daniel J. Huey, Joel A. Hughes, Jane M. Griffith Univ Australian Rivers Inst Nathan Qld 4111 Australia Western Australian Museum Terr Zool & Mol Systemat Unit 49 Kew St Welshpool WA 6986 Australia

The utility of restriction-site associated DNA sequencing (RADseq) to resolve fine-scale population structure was tested on an abundant and vagile fish species in a tropical river. Australia's most widespread freshwater fish, the "extreme disperser" Leiopotherapon unicolor was sampled from 6 locations in an unregulated system, the Daly River in Australia's Northern Territory. Despite an expectation of high connectivity based on life history knowledge of this species derived from arid zone habitats, L. unicolor was not a panmictic population in the tropical lower Daly. Using similar to 14 000 polymorphic RADseq loci, we found a pattern of upstream versus downstream population subdivision and evidence for differentiation among tributary populations. The magnitude of population structure was low with narrow confidence intervals (global F-ST = 0.014;95% CI = 0.012-0.016). Confidence intervals around pairwise F-ST estimates were all nonzero and consistent with the results of clustering analyses. This population structure was not explained by spatially heterogeneous selection acting on a subset of loci, or by sampling groups of closely related individuals (average within-site relatedness approximate to 0). One implication of the low but significant structure observed in the tropics is the possibility that L. unicolor may exhibit contrasting patterns of migratory biology in tropical versus arid zone habitats. We conclude that the RADseq revolution holds promise for delineating subtle patterns of population subdivision in species characterized by high within-population variation and low among-population differentiation.

关键词： DAPC F-ST outlier outflank RADseq sNMF sparse non-negative matrix factorization

来源：评论

学校读者我要写书评

暂无评论

An Efficient Approach for Fault Detection, Isolation, and Data Recovery of Self-Validating Multifunctional Sensors

引用

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 2017年第3期66卷 543-558页

作者： Yang, Jingli Chen, Yinsheng Zhang, Lili Harbin Inst Technol Sch Elect Engn & Automat Harbin 150001 Peoples R China Harbin Univ Commerce Coll Basic Sci Harbin 150001 Peoples R China

A novel fault detection, isolation, and data recovery (FDIR) approach for self-validating multifunctional sensors is presented in this paper. To improve the fault detection accuracy under multiple steady conditions for multifunctional sensors, a sparse non-negative matrix factorization (SNMF)-based model is employed to accomplish fault detection through a combination of newly proposed C-2 and squared prediction error (SPE) statistics. Furthermore, a self-adaptive multiple-variable reconstruction strategy (SMVR) is proposed to achieve high accuracy on multiple-fault isolation and data recovery for faulty sensitive units. The performance of the proposed approach is fully verified in a real experimental system for self-validating multifunctional sensors, and it is compared with those of other fault detection models, such as principal component analysis (PCA), non-negative matrix factorization (NMF), and fault isolation algorithms, such as PCA-based contribution plots and SNMFbased contribution plots. The experimental results demonstrate that the proposed approach provides an excellent solution to the FDIR of self-validating multifunctional sensors.

关键词： Data recovery fault detection and isolation self-adaptive multiple-variable reconstruction self-validating multifunctional sensors sparse non-negative matrix factorization

来源：评论

学校读者我要写书评

暂无评论

Learning Part-based Dictionary by sparse NMF for Face Gender Recognition 8

Learning Part-based Dictionary by Sparse NMF for Face Gender...

引用

8th International Symposium on Computational Intelligence and Design (ISCID)

作者： Ge, Jiali Zhou, Tong Zhang, Fan Tse, Ken Nanjing Tech Univ Sch Control Sci & Elect Engn Nanjing Jiangsu Peoples R China Xidian Univ Dept Comp Sci & Technol Xian Peoples R China Xian Univ Technol Sch Automat & Informat Engn Xian Peoples R China Univ Calif Los Angeles Dept Comp Sci Los Angeles CA 90024 USA

ISBN: (纸本)9781467395878

Face gender recognition is a very challenging problem in computer vision, which plays an important role in many visual applications. In this paper, we present a framework that combines the unsupervised dictionary learning and supervised classifier training together to this gender recognition problem. We firstly apply sparse non-negative matrix factorization (sparse NMF) to learn intrinsic part-based dictionary from face images in an unsupervised manner. After that we encode all the data by the learned dictionary, and train a SVM or logistic regression classifier in a supervised manner on those representations. Our experimental results show that the learned dictionaries by sparse NMF can not only capture meaningful features from the faces, but also boost the performance of the subsequent classifier in terms of classification accuracies and speeds.

关键词： face gender reconition sparse non-negative matrix factorization SVM logistic regression dictionary learning

来源：评论

学校读者我要写书评

暂无评论

Learning Part-based Dictionary by sparse NMF for Face Gender Recognition

Learning Part-based Dictionary by Sparse NMF for Face Gender...

引用

International Symposium on Computational Intelligence and Design

作者： Jiali Ge Tong Zhou Fan Zhang Ken Tse School of Control Science and Electrical Engineering Nanjing Tech University Nanjing China Department of Computer Science and Technology Xidian University Xi'an China School of Automation and Information Engineering Xi'an University Of Technology Xi'an China Department of Computer Science University of California Los Angeles USA

ISBN: (纸本)9781467395885

Face gender recognition is a very challenging problem in computer vision, which plays an important role in many visual applications. In this paper, we present a framework that combines the unsupervised dictionary learning and supervised classifier training together to this gender recognition problem. We firstly apply sparse nonnegative matrix factorization (sparse NMF) to learn intrinsic part-based dictionary from face images in an unsupervised manner. After that we encode all the data by the learned dictionary, and train a SVM or logistic regression classifier in a supervised manner on those representations. Our experimental results show that the learned dictionaries by sparse NMF can not only capture meaningful features from the faces, but also boost the performance of the subsequent classifier in terms of classification accuracies and speeds.

关键词： face gender reconition sparse non-negative matrix factorization SVM logistic regression dictionary learning

来源：评论

学校读者我要写书评

暂无评论

Spatio-spectral fusion of satellite images based on dictionary-pair learning

引用

INFORMATION FUSION 2014年第0期18卷 148-160页

作者： Song, Huihui Huang, Bo Zhang, Kaihua Zhang, Hankui Chinese Univ Hong Kong Dept Geog & Resource Management Hong Kong Hong Kong Peoples R China Hong Kong Polytech Univ Dept Comp Hong Kong Hong Kong Peoples R China

This paper proposes a novel spatial and spectral fusion method for satellite multispectral and hyperspectral (or high-spectral) images based on dictionary-pair learning. By combining the spectral information from sensors with low spatial resolution but high spectral resolution (LSHS) and the spatial information from sensors with high spatial resolution but low spectral resolution (HSLS), this method aims to generate fused data with both high spatial and spectral resolution. Based on the sparse non-negative matrix factorization technique, this method first extracts spectral bases of LSHS and HSLS images by making full use of the rich spectral information in LSHS data. The spectral bases of these two categories data then formulate a dictionary-pair due to their correspondence in representing each pixel spectra of LSHS data and HSLS data, respectively. Subsequently, the LSHS image is spatial unmixed by representing the HSLS image with respect to the corresponding learned dictionary to derive its representation coefficients. Combining the spectral bases of LSHS data and the representation coefficients of HSLS data, fused data are finally derived which are characterized by the spectral resolution of LSHS data and the spatial resolution of HSLS data. The experiments are carried out by comparing the proposed method with two representative methods on both simulation data and actual satellite images, including the fusion of Landsat/ETM+ and Aqua/MODIS data and the fusion of EO-1/Hyperion and SPOT5/HRG multispectral images. By visually comparing the fusion results and quantitatively evaluating them in term of several measurement indices, it can be concluded that the proposed method is effective in preserving both the spectral information and spatial details and performs better than the comparison approaches. (C) 2013 Elsevier B.V. All rights reserved.

关键词： Spatio-spectral fusion High spatial resolution High spectral resolution Dictionary-pair learning sparse non-negative matrix factorization

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

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

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：