8,219 Hits in 2.5 sec

Joint Weakly and Semi-Supervised Deep Learning for Localization and Classification of Masses in Breast Ultrasound Images

Seung Yeon Shin, Soochahn Lee, Il Dong Yun, Sun Mi Kim, Kyoung Mu Lee
2019 IEEE Transactions on Medical Imaging  
We propose a framework for localization and classification of masses in breast ultrasound images.  ...  We propose a systematic weakly and semi-supervised training scenario with appropriate training loss selection.  ...  This method has been developed in a typical medical image setting, where strong annotations are available for only a portion of available medical image data due to the limited resources of physicians.  ... 
doi:10.1109/tmi.2018.2872031 pmid:30273145 fatcat:57g2ijygvrexzovwsurqlt2jni

Weakly Supervised 3D Classification of Chest CT using Aggregated Multi-Resolution Deep Segmentation Features [article]

Anindo Saha, Fakrul I. Tushar, Khrystyna Faryna, Vincent M. D'Anniballe, Rui Hou, Maciej A. Mazurowski, Geoffrey D. Rubin, Joseph Y. Lo
2020 arXiv   pre-print
Weakly supervised disease classification of CT imaging suffers from poor localization owing to case-level annotations, where even a positive scan can hold hundreds to thousands of negative slices along  ...  In this research, we propose a medical classifier that leverages the semantic structural concepts learned via multi-resolution segmentation feature maps, to guide weakly supervised 3D classification of  ...  ACKNOWLEDGEMENTS This work was supported in part by developmental funds of the Duke Cancer Institute as part of the NIH/NCI P30 CA014236 Cancer Center Support Grant.  ... 
arXiv:2011.00149v1 fatcat:fy4spbmhc5g4vnqy3powsxrrl4

Predicting Semantic Descriptions from Medical Images with Convolutional Neural Networks [chapter]

Thomas Schlegl, Sebastian M. Waldstein, Wolf-Dieter Vogl, Ursula Schmidt-Erfurth, Georg Langs
2015 Lecture Notes in Computer Science  
Weakly supervised learning approaches can link volume-level labels to image content but suffer from the typical label distributions in medical imaging data where only a small part consists of clinically  ...  Learning representative computational models from medical imaging data requires large training data sets. Often, voxel-level annotation is unfeasible for sufficient amounts of data.  ...  The corresponding situation in medical imaging consists of information that somewhere in the image there is a certain abnormality. Weakly supervised approaches learn from these weak or noisy labels.  ... 
doi:10.1007/978-3-319-19992-4_34 fatcat:f7akmhypuzgjjael77donync7e

Deep learning with mixed supervision for brain tumor segmentation

Pawel Mlynarski, Hervé Delingette, Antonio Criminisi, Nicholas Ayache
2019 Journal of Medical Imaging  
The idea of our approach is to extend segmentation networks with an additional branch performing image-level classification.  ...  This type of training data is particularly costly, as manual delineation of tumors is not only time-consuming but also requires medical expertise.  ...  Another type of weakly-supervised methods aims to detect objects in natural images based on classification of image subregions [19, 20] using pretrained classification networks such as VGG-Net [21]  ... 
doi:10.1117/1.jmi.6.3.034002 pmid:31423456 pmcid:PMC6689144 fatcat:nndbxyudwrhpfo2ewyf63fz7j4

Towards to Reasonable Decision Basis in Automatic Bone X-Ray Image Classification: A Weakly-Supervised Approach

Jianjie Lu, Kai-yu Tong
A weakly-supervised framework is proposed that cannot only make class inference but also provides reasonable decision basis in bone X-ray images.  ...  model to make classification prediction from the activation areas; (3) label lesions in very few images and guide the model to learn simultaneously.  ...  Introduction Deep learning, particularly convolutional network (CNN), has achieved high accuracies in classification of medical images (Litjens et al. 2017) .  ... 
doi:10.1609/aaai.v33i01.33019985 fatcat:2qzhqekilvfdll4nyz2xzso2qm

Deep Learning with Mixed Supervision for Brain Tumor Segmentation [article]

Pawel Mlynarski, Hervé Delingette, Antonio Criminisi, Nicholas Ayache
2018 arXiv   pre-print
The idea of our approach is to extend segmentation networks with an additional branch performing image-level classification.  ...  The model is jointly trained for segmentation and classification tasks in order to exploit information contained in weakly-annotated images while preventing the network to learn features which are irrelevant  ...  Another type of weakly-supervised methods aims to detect objects in natural images based on classification of image subregions [19, 20] using pretrained classification networks such as VGG-Net [21]  ... 
arXiv:1812.04571v1 fatcat:hbwfgqkunjdkdbhupxgyksy5fm

Weakly and Semi Supervised Detection in Medical Imaging via Deep Dual Branch Net [article]

Ran Bakalo, Jacob Goldberger, Rami Ben-Ari
2020 arXiv   pre-print
This study presents a novel deep learning architecture for multi-class classification and localization of abnormalities in medical imaging illustrated through experiments on mammograms.  ...  Our method enables detection of abnormalities at full mammogram resolution for both weakly and semi-supervised settings.  ...  analysis of the classification results of our weakly supervised method, and multiclass classification results.  ... 
arXiv:1904.12589v3 fatcat:w7bx5endajaofcdfeyap3kfhle

CheXseg: Combining Expert Annotations with DNN-generated Saliency Maps for X-ray Segmentation [article]

Soham Gadgil, Mark Endo, Emily Wen, Andrew Y. Ng, Pranav Rajpurkar
2021 arXiv   pre-print
Medical image segmentation models are typically supervised by expert annotations at the pixel-level, which can be expensive to acquire.  ...  We find that CheXseg improves upon the performance (mIoU) of fully-supervised methods that use only pixel-level expert annotations by 9.7% and weakly-supervised methods that use only DNN-generated saliency  ...  Saliency maps are a popular set of explanation methods that highlight regions of the image that are important for disease classification, but they have been shown to be untrustworthy for medical image  ... 
arXiv:2102.10484v2 fatcat:p5bjcfdg7bbkvmrecm3p6md5em

Deep learning of feature representation with multiple instance learning for medical image analysis

Yan Xu, Tao Mo, Qiwei Feng, Peilin Zhong, Maode Lai, Eric I-Chao Chang
2014 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
This paper studies the effectiveness of accomplishing high-level tasks with a minimum of manual annotation and good feature representations for medical images.  ...  In medical image analysis, objects like cells are characterized by significant clinical features.  ...  Settings We studied four different types of features on all 200x200 patches, and finished classification of fully supervised learning and of weakly supervised learning on them.  ... 
doi:10.1109/icassp.2014.6853873 dblp:conf/icassp/XuMFZLC14 fatcat:wrctt7kxlvhotoipewsw3nqmbm

Weakly-Supervised Segmentation for Disease Localization in Chest X-Ray Images [article]

Ostap Viniavskyi, Mariia Dobko, Oles Dobosevych
2020 arXiv   pre-print
In this paper, we propose a novel approach to the semantic segmentation of medical chest X-ray images with only image-level class labels as supervision.  ...  First, we generate pseudo segmentation labels of abnormal regions in the training images through a supervised classification model enhanced with a regularization procedure.  ...  Acknowledgements This research was supported by SoftServe and Faculty of Applied Sciences at Ukrainian Catholic University (UCU), whose collaboration allowed to create SoftServe Research Group at UCU.  ... 
arXiv:2007.00748v1 fatcat:bjtbj6vvdzbyndbfwy5r735oxi

Medical image segmentation with imperfect 3D bounding boxes [article]

Ekaterina Redekop, Alexey Chernyavskiy
2021 arXiv   pre-print
While current weakly-supervised approaches that use 2D bounding boxes as weak labels can be applied to medical image segmentation, we show that their success is limited in cases when the assumption about  ...  The development of high quality medical image segmentation algorithms depends on the availability of large datasets with pixel-level labels.  ...  First, we show the limitation of current weakly-supervised approaches that use 2D bounding boxes as weak labels, when applied to medical image segmentation in cases when the bounding boxes are not tight  ... 
arXiv:2108.03300v1 fatcat:an4te3vgnncbdkf3xul336zqom

Self-Transfer Learning for Fully Weakly Supervised Object Localization [article]

Sangheum Hwang, Hyo-Eun Kim
2016 arXiv   pre-print
We evaluate the proposed STL framework using two medical image datasets, chest X-rays and mammograms, and achieve signiticantly better localization performance compared to previous weakly supervised approaches  ...  Thus a weakly supervised framework for object localization is introduced. The term "weakly" means that this framework only uses image-level labeled datasets to train a network.  ...  Weakly supervised localization based on CNN For the task of image classification, CNN works well by virtue of its ability to extract useful features which discriminate the classes.  ... 
arXiv:1602.01625v1 fatcat:kp77sese3jdj3fkdrpfssqvjou

Iterative augmentation of visual evidence for weakly-supervised lesion localization in deep interpretability frameworks: application to color fundus images

Cristina Gonzalez-Gonzalo, Bart Liefers, Bram van Ginneken, Clara I. Sanchez
2020 IEEE Transactions on Medical Imaging  
of weakly-supervised localization of different types of DR and AMD abnormalities, both qualitatively and quantitatively.  ...  We propose a deep visualization method to generate interpretability of DL classification tasks in medical imaging by means of visual evidence augmentation.  ...  weakly-supervised detection of retinal lesions.  ... 
doi:10.1109/tmi.2020.2994463 pmid:32746093 fatcat:enrasntk5nagffq2hu5ceqbo2q

Manifold-driven Attention Maps for Weakly Supervised Segmentation [article]

Sukesh Adiga V, Jose Dolz, Herve Lombaert
2020 arXiv   pre-print
To mitigate this problem, weakly supervised learning has emerged as an efficient alternative, which employs image-level labels, scribbles, points, or bounding boxes as supervision.  ...  Segmentation using deep learning has shown promising directions in medical imaging as it aids in the analysis and diagnosis of diseases.  ...  of a GPU.  ... 
arXiv:2004.03046v1 fatcat:cla7kqcdojbspbl6igue2uj4gy

Deconvolutional Feature Stacking for Weakly-Supervised Semantic Segmentation [article]

Hyo-Eun Kim, Sangheum Hwang
2016 arXiv   pre-print
Under the weak supervision (image-level labels), the proposed framework shows promising results on lesion segmentation in medical images (chest X-rays) and achieves state-of-the-art performance on the  ...  A weakly-supervised semantic segmentation framework with a tied deconvolutional neural network is presented.  ...  In a binary classification problem (e.g., abnormality detection in medical images), K is two as usual. For the multi-label classification, K includes an additional class for 'background'.  ... 
arXiv:1602.04984v3 fatcat:jjcekj7fjvfw7amrb42tke2lpu
« Previous Showing results 1 — 15 out of 8,219 results