Deep Reinforcement Learning for Active Breast Lesion Detection from DCE-MRI

Maicas, Gabriel; Carneiro, Gustavo; Bradley, Andrew P.; Nascimento, Jacinto C.; Reid, Ian

doi:10.1007/978-3-319-66179-7_76

Cited by 70 publications

(79 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared with traditional methods, the main advantage of deep learning models is its capability in extracting highly representative features in a data‐driven way. Recently, the efficacy of deep neural networks has been evaluated in breast cancer classification tasks . However, these works either used a small‐size dataset or needed manual annotations on lesions during the training phase.…”

Section: Discussionmentioning

confidence: 99%

“…Recently, the efficacy of deep neural networks has been evaluated in breast cancer classification tasks. [28][29][30] However, these works either used a small-size dataset or needed manual annotations on lesions during the training phase. Directly localizing breast cancers in 3D radiology images with only image-level supervision has not yet been extensively explored.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Weakly supervised 3D deep learning for breast cancer classification and localization of the lesions in MR images

Zhou

Luo

Dou

et al. 2019

Magnetic Resonance Imaging

112

View full text Add to dashboard Cite

Background The usefulness of 3D deep learning‐based classification of breast cancer and malignancy localization from MRI has been reported. This work can potentially be very useful in the clinical domain and aid radiologists in breast cancer diagnosis. Purpose To evaluate the efficacy of 3D deep convolutional neural network (CNN) for diagnosing breast cancer and localizing the lesions at dynamic contrast enhanced (DCE) MRI data in a weakly supervised manner. Study Type Retrospective study. Subjects A total of 1537 female study cases (mean age 47.5 years ±11.8) were collected from March 2013 to December 2016. All the cases had labels of the pathology results as well as BI‐RADS categories assessed by radiologists. Field Strength/Sequence 1.5 T dynamic contrast‐enhanced MRI. Assessment Deep 3D densely connected networks were trained under image‐level supervision to automatically classify the images and localize the lesions. The dataset was randomly divided into training (1073), validation (157), and testing (307) subsets. Statistical Tests Accuracy, sensitivity, specificity, area under receiver operating characteristic curve (ROC), and the McNemar test for breast cancer classification. Dice similarity for breast cancer localization. Results The final algorithm performance for breast cancer diagnosis showed 83.7% (257 out of 307) accuracy (95% confidence interval [CI]: 79.1%, 87.4%), 90.8% (187 out of 206) sensitivity (95% CI: 80.6%, 94.1%), 69.3% (70 out of 101) specificity (95% CI: 59.7%, 77.5%), with the area under the curve ROC of 0.859. The weakly supervised cancer detection showed an overall Dice distance of 0.501 ± 0.274. Data Conclusion 3D CNNs demonstrated high accuracy for diagnosing breast cancer. The weakly supervised learning method showed promise for localizing lesions in volumetric radiology images with only image‐level labels. Level of Evidence: 4 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019;50:1144–1151.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Weakly supervised 3D deep learning for breast cancer classification and localization of the lesions in MR images

Zhou

Luo

Dou

et al. 2019

Magnetic Resonance Imaging

112

View full text Add to dashboard Cite

show abstract

“…With a well-qualified exploration space, the DQN is promised to capture the prior about pancreas location even with limited data. Previous work [24]- [26] has proposed DQN for efficient object detection, and recent work [27], [28] applied it to anatomical landmark and breast lesion detection, suggesting its competence in MIA. Nonetheless, the incorporation of DQN into more complex Medical Image Segmentation problems has never been expolred, which is the focus of our work.…”

Section: Object Detectionmentioning

confidence: 99%

Deep Q Learning Driven CT Pancreas Segmentation With Geometry-Aware U-Net

Man

Huang

Feng

et al. 2019

IEEE Trans. Med. Imaging

131

View full text Add to dashboard Cite

Segmentation of pancreas is important for medical image analysis, yet it faces great challenges of class imbalance, background distractions and non-rigid geometrical features. To address these difficulties, we introduce a Deep Q Network(DQN) driven approach with deformable U-Net to accurately segment the pancreas by explicitly interacting with contextual information and extract anisotropic features from pancreas. The DQN based model learns a context-adaptive localization policy to produce a visually tightened and precise localization bounding box of the pancreas. Furthermore, deformable U-Net captures geometryaware information of pancreas by learning geometrically deformable filters for feature extraction. Experiments on NIH dataset validate the effectiveness of the proposed framework in pancreas segmentation.

show abstract

“…They use deep Q-learning for class-specific object detection which allows an agent to stepwise deform a bounding box in its size, position and aspect ratio to fit an object. A similar algorithm was proposed by [8], which allowed for three-dimensional bounding box transformations in size and position to detect breast lesions. This approach reduced the runtime of breast lesion detection without reducing the detection accuracy compared to other approaches in the field.…”

Section: Related Workmentioning

confidence: 99%

Multi-stage Reinforcement Learning for Object Detection

König

Malberg

Martens

et al. 2019

Advances in Intelligent Systems and Computing

View full text Add to dashboard Cite

We present a reinforcement learning approach for detecting objects within an image. Our approach performs a step-wise deformation of a bounding box with the goal of tightly framing the object. It uses a hierarchical tree-like representation of predefined region candidates, which the agent can zoom in on. This reduces the number of region candidates that must be evaluated so that the agent can afford to compute new feature maps before each step to enhance detection quality. We compare an approach that is based purely on zoom actions with one that is extended by a second refinement stage to fine-tune the bounding box after each zoom step. We also improve the fitting ability by allowing for different aspect ratios of the bounding box. Finally, we propose different reward functions to lead to a better guidance of the agent while following its search trajectories. Experiments indicate that each of these extensions leads to more correct detections. The best performing approach comprises a zoom stage and a refinement stage, uses aspect-ratio modifying actions and is trained using a combination of three different reward metrics.

show abstract

Deep Reinforcement Learning for Active Breast Lesion Detection from DCE-MRI

Cited by 70 publications

References 15 publications

Weakly supervised 3D deep learning for breast cancer classification and localization of the lesions in MR images

Weakly supervised 3D deep learning for breast cancer classification and localization of the lesions in MR images

Deep Q Learning Driven CT Pancreas Segmentation With Geometry-Aware U-Net

Multi-stage Reinforcement Learning for Object Detection

Contact Info

Product

Resources

About