-
公开(公告)号：US20220262105A1
公开(公告)日：2022-08-18
申请号：US17625313
申请日：2020-07-17
申请人： Zongwei ZHOU , Vatsal SODHA , Md, Mahfuzur RAHMAN SIDDIQUEE , Ruibin FENG , Nima TAJBAKHSH , Jianming LIANG , Arizona Board of Regents on behalf of Arizona State University
发明人： Zongwei Zhou , Vatsal Sodha , Md Mahfuzur Rahman Siddiquee , Ruibin Feng , Nima Tajbakhsh , Jianming Liang
IPC分类号： G06V10/774 , G06V10/82 , G06V10/98 , G06V10/776
摘要： Described herein are means for generating source models for transfer learning to application specific models used in the processing of medical imaging. In some embodiments, the method comprises: identifying a group of training samples, wherein each training sample in the group of training samples includes an image; for each training sample in the group of training samples: identifying an original patch of the image corresponding to the training sample; identifying one or more transformations to be applied to the original patch; generating a transformed patch by applying the one or more transformations to the identified patch; and training an encoder-decoder network using a group of transformed patches corresponding to the group of training samples, wherein the encoder-decoder network is trained to generate an approximation of the original patch from a corresponding transformed patch, and wherein the encoder-decoder network is trained to minimize a loss function that indicates a difference between the generated approximation of the original patch and the original patch. The source models significantly enhance the transfer learning performance for many medical imaging tasks including, but not limited to, disease/organ detection, classification, and segmentation. Other related embodiments are disclosed.
-
公开(公告)号：US20200380695A1
公开(公告)日：2020-12-03
申请号：US16885579
申请日：2020-05-28
摘要： Methods, systems, and media for segmenting images are provided. In some embodiments, the method comprises: generating an aggregate U-Net comprised of a plurality of U-Nets, wherein each U-Net in the plurality of U-Nets has a different depth, wherein each U-Net is comprised of a plurality of nodes Xi,j, wherein i indicates a down-sampling layer the U-Net, and wherein j indicates a convolution layer of the U-Net; training the aggregate U-Net by: for each training sample in a group of training samples, calculating, for each node in the plurality of nodes Xi,j, a feature map xi,j, wherein xi,j is based on a convolution operation performed on a down-sampling of an output from Xi−1,j when j=0, and wherein xi,j is based on a convolution operation performed on an up-sampling operation of an output from Xi+1,j−1 when j>0; and predicting a segmentation of a test image using the trained aggregate U-Net.
-
公开(公告)号：US12094190B2
公开(公告)日：2024-09-17
申请号：US17675929
申请日：2022-02-18
发明人： Diksha Goyal , Jianming Liang
IPC分类号： G06V10/778 , G06T7/00 , G06T7/11 , G06T7/194 , G06V10/774 , G06V10/82
CPC分类号： G06V10/7788 , G06T7/0012 , G06T7/11 , G06T7/194 , G06V10/7747 , G06V10/82 , G06T2200/24 , G06T2207/20081 , G06T2207/20084 , G06T2207/20092 , G06T2207/30004
摘要： Medical image segmentation using interactive refinement, in which the trained deep models are then utilized for the processing of medical imaging are described. Operating a two-step deep learning training framework including receiving original input images at the deep learning training framework; generating an initial prediction image specifying image segmentation by base segmentation model; receiving user input guidance signals; routing each of (i) the original input images, (ii) the initial prediction image, and (iii) the user input guidance signals to an InterCNN; generating a refined prediction image specifying refined image segmentation by processing each of the (i) the original input images, (ii) the initial prediction image, and (iii) the user input guidance signals through the InterCNN to render the refined prediction image incorporating the user input guidance signals; and outputting a refined segmentation mask to the deep learning training framework as a guidance signal.
-
公开(公告)号：US11922628B2
公开(公告)日：2024-03-05
申请号：US17224886
申请日：2021-04-07
发明人： Zongwei Zhou , Vatsal Sodha , Jiaxuan Pang , Jianming Liang
IPC分类号： G06V10/00 , G06F18/21 , G06F18/214 , G06N3/045 , G06N3/088 , G06T3/00 , G06T7/11 , G06V10/26 , G06V10/77 , G16H30/40
CPC分类号： G06T7/11 , G06F18/2155 , G06F18/2163 , G06N3/045 , G06N3/088 , G06T3/00 , G06V10/26 , G06V10/7715 , G16H30/40 , G06V2201/03
摘要： Described herein are means for generation of self-taught generic models, named Models Genesis, without requiring any manual labeling, in which the Models Genesis are then utilized for the processing of medical imaging. For instance, an exemplary system is specially configured for learning general-purpose image representations by recovering original sub-volumes of 3D input images from transformed 3D images. Such a system operates by cropping a sub-volume from each 3D input image; performing image transformations upon each of the sub-volumes cropped from the 3D input images to generate transformed sub-volumes; and training an encoder-decoder architecture with skip connections to learn a common image representation by restoring the original sub-volumes cropped from the 3D input images from the transformed sub-volumes generated via the image transformations. A pre-trained 3D generic model is thus provided, based on the trained encoder-decoder architecture having learned the common image representation which is capable of identifying anatomical patterns in never before seen 3D medical images having no labeling and no annotation. More importantly, the pre-trained generic models lead to improved performance in multiple target tasks, effective across diseases, organs, datasets, and modalities.
-
公开(公告)号：US20220270357A1
公开(公告)日：2022-08-25
申请号：US17675929
申请日：2022-02-18
发明人： Diksha Goyal , Jianming Liang
IPC分类号： G06V10/778 , G06T7/194 , G06V10/82 , G06V10/774 , G06T7/00 , G06T7/11
摘要： Described herein are means for implementing medical image segmentation using interactive refinement, in which the trained deep models are then utilized for the processing of medical imaging. For instance, an exemplary system is specially configured for operating a two-step deep learning training framework including means for receiving original input images at the deep learning training framework; means for generating an initial prediction image specifying image segmentation by processing the original input images through the base segmentation model to render the initial prediction image in the absence of user input guidance signals; means for receiving user input guidance signals indicating user-guided segmentation refinements to the initial prediction image; means for routing each of (i) the original input images, (ii) the initial prediction image, and (iii) the user input guidance signals to an InterCNN; means for generating a refined prediction image specifying refined image segmentation by processing each of the (i) the original input images, (ii) the initial prediction image, and (iii) the user input guidance signals through the InterCNN to render the refined prediction image incorporating the user input guidance signals; and means for outputting a refined segmentation mask based on application of the user input guidance signals to the deep learning training framework as a guidance signal. Other related embodiments are disclosed.
-
公开(公告)号：US11328430B2
公开(公告)日：2022-05-10
申请号：US16885579
申请日：2020-05-28
摘要： Methods, systems, and media for segmenting images are provided. In some embodiments, the method comprises: generating an aggregate U-Net comprised of a plurality of U-Nets, wherein each U-Net in the plurality of U-Nets has a different depth, wherein each U-Net is comprised of a plurality of nodes Xi,j, wherein i indicates a down-sampling layer the U-Net, and wherein j indicates a convolution layer of the U-Net; training the aggregate U-Net by: for each training sample in a group of training samples, calculating, for each node in the plurality of nodes Xi,j, a feature map xi,j, wherein xi,j is based on a convolution operation performed on a down-sampling of an output from Xi−1,j when j=0, and wherein xi,j is based on a convolution operation performed on an up-sampling operation of an output from Xi+1,j−1 when j>0; and predicting a segmentation of a test image using the trained aggregate U-Net.
-
公开(公告)号：US20220114733A1
公开(公告)日：2022-04-14
申请号：US17497528
申请日：2021-10-08
发明人： Ruibin Feng , Zongwei Zhou , Jianming Liang
摘要： Described herein are means for implementing contrastive learning via reconstruction within a self-supervised learning framework, in which the trained deep models are then utilized for the processing of medical imaging. For instance, an exemplary system is specially configured for performing a random cropping operation to crop a 3D cube from each of a plurality of medical images received at the system as input; performing a resize operation of the cropped 3D cubes; performing an image reconstruction operation of the resized and cropped 3D cubes to predict the whole image represented by the original medical images received; and generating a reconstructed image which is analyzed for reconstruction loss against the original image representing a known ground truth image to the reconstruction loss function. Other related embodiments are disclosed.
-
公开(公告)号：US20210150710A1
公开(公告)日：2021-05-20
申请号：US17098422
申请日：2020-11-15
摘要： Not only is annotating medical images tedious and time consuming, but it also demands costly, specialty-oriented expertise, which is not easily accessible. To address this challenge, a new self-supervised framework is introduced: TransVW (transferable visual words), exploiting the prowess of transfer learning with convolutional neural networks and the unsupervised nature of visual word extraction with bags of visual words, resulting in an annotation-efficient solution to medical image analysis. TransVW was evaluated using NIH ChestX-ray14 to demonstrate its annotation efficiency. When compared with training from scratch and ImageNet-based transfer learning, TransVW reduces the annotation efforts by 75% and 12%, respectively, in addition to significantly accelerating the convergence speed. More importantly, TransVW sets new records: achieving the best average AUC on all 14 diseases, the best individual AUC scores on 10 diseases, and the second best individual AUC scores on 3 diseases. This performance is unprecedented, because heretofore no self-supervised learning method has outperformed ImageNet-based transfer learning and no annotation reduction has been reported for self-supervised learning. These achievements are contributable to a simple yet powerful observation: The complex and recurring anatomical structures in medical images are natural visual words, which can be automatically extracted, serving as strong yet free supervision signals for CNNs to learn generalizable and transferable image representation via self-supervision.
-
公开(公告)号：US20240339200A1
公开(公告)日：2024-10-10
申请号：US18627831
申请日：2024-04-05
发明人： DongAo Ma , Jiaxuan Pang , Jianming Liang
IPC分类号： G16H30/40 , G06V10/764 , G16H30/20
CPC分类号： G16H30/40 , G06V10/764 , G16H30/20
摘要： Exemplary systems include means for receiving medical image data at the system from a plurality of datasets provided via publicly available sources; evaluating the medical image data for the presence of expert notation embedded within the medical image data; determining the expert notations embedded within the medical image data are formatted using inconsistent and heterogeneous labeling across the plurality of datasets; generating an interim AI model by applying a task head classifier to learn the annotations of the expert notations embedded within the medical image data to generate an interim AI model; scaling the interim AI model having the learned annotations of the expert notations embedded therein to additional tasks by applying multi-task heads using cyclical pre-training of the interim AI model trained previously to generate task-specific AI models, with each respective task-specific AI model having differently configured task-specific learning objectives; training a pre-trained AI model specially configured for an application-specific target task by applying task re-visitation training forcing the pre-trained AI model being trained to re-visit all tasks in each round of training and forcing the pre-trained AI model being trained to re-use all accrued knowledge to improve learning by the pre-trained AI model being trained against the current application-specific target task for which the pre-trained AI model is being trained.
-
公开(公告)号：US20240078434A1
公开(公告)日：2024-03-07
申请号：US18241811
申请日：2023-09-01
发明人： Zuwei Guo , Nahid Ul Islam , Jianming Liang
IPC分类号： G06N3/0895 , G06N3/0455 , G06T7/00 , G06V10/774 , G06V10/82 , G06V20/50 , G06V20/70
CPC分类号： G06N3/0895 , G06N3/0455 , G06T7/0012 , G06V10/774 , G06V10/82 , G06V20/50 , G06V20/70 , G06T2207/20081 , G06T2207/20084 , G06T2207/30004 , G06V2201/03
摘要： The system receives a plurality of medical images and integrates Self-Supervised machine Learning (SSL) instructions for performing a discriminative learning operation, a restorative learning operation, and an adversarial learning operation into a model for processing the received plurality of medical images. The model is configured with each of a discriminative encoder, a restorative decoder, and an adversarial encoder. Each of the discriminative encoder and the restorative decoder are configured to be skip connected, forming an encoder-decoder. Step-wise incremental training to incrementally train each of the discriminative encoder, the restorative decoder, and the adversarial encoder is performed, in particular: pre-training the discriminative encoder via discriminative learning; attaching the pre-trained discriminative encoder to the restorative decoder to configure the encoder-decoder as a pre-trained encoder-decoder; and training the pre-trained encoder-decoder of the model using joint discriminative and restorative learning. The pre-trained encoder-decoder is associated with the adversarial encoder. The pre-trained encoder-decoder associated with the adversarial encoder is trained through discriminative, restorative, and adversarial learning to render a trained model for the processing of the received plurality of medical images. The plurality of medical images are processed through the model using the trained model.
-
-
-
-
-
-
-
-
-
搜索结果
91 条记录 (耗时 0.052 秒)
