The combination of digital breast tomosynthesis (DBT) with other imaging modalities has been investigated in order to improve the detection and diagnosis of breast cancer. Mechanical Imaging (MI) measures the stress over the surface of the compressed breast, using a pressure sensor, during radiographic examination and its response has shown a correlation with the presence of malignant lesions. Thus, the combination of DBT and MI (DBTMI) has shown potential to reduce false positive results in breast cancer screening. However, compared to the conventional DBT exam, the presence of the MI sensor during mammographic image acquisition may cause a slight increase in the radiation dose. This work presents a proposal to reduce the radiation dose in DBTMI exams by removing some projections from the original set and replacing them with synthetic projections generated by a video frame interpolation (VFI) neural network. We compared several DBTMI acquisition arrangements, considering the removal of 16% of the original projections, using a deformable physical breast phantom, and evaluated the quality of the reconstructed images based on the Normalized Root Mean Squared Error (NRMSE). The results showed that, for some arrangements, the slices reconstructed with the addition of synthetic DBTMI projections presented better quality than when they were reconstructed with the reduced set of projections. Further studies must be carried out to optimize the interpolation approach.
KEYWORDS: Digital breast tomosynthesis, Image quality, Video, Breast, Visualization, Neural networks, Reconstruction algorithms, Signal to noise ratio, Medical imaging, Image analysis
The angular range and number of projections are parameters that directly influence the image quality and the visibility of lesions in digital breast tomosynthesis (DBT). The medical field is taking advantage of the increasing performance of machine learning algorithms with the use of complex data-driven models, known as deep learning (DL) networks. The use of DL has also been highlighted in the tasks of video frame interpolation (VFI) for the synthesis of new images in order to increase the frame rate per second. In the present work, we use a residual refinement interpolation network (RRIN) to generate new synthetic DBT projections from pairs of real projections. We studied two different approaches: first, we increased the number of projections before reconstruction using the synthetic images, with the aim of improving the quality of the reconstructed slices without increasing the radiation dose to the patient. In the second, we investigated the effect of replacing existing projections with synthetic ones, with the objective of reducing the radiation dose and acquisition time. In the first approach, we used virtual phantoms to generate sets of DBT projections to train the network. We then evaluated the contrast-to-noise ratio (CNR) of simulated microcalcifications after reconstruction. The CNR was higher for all sets where supplementary images were added compared to those with only real images. In the second approach, we trained the network with clinical data and tested it with images acquired with a physical anthropomorphic breast phantom. Both the projections and the slices showed good similarity with the real ones, suggesting that the use of VFI networks to generate DBT projections is promising. However, further studies should be carried out to assess the feasibility of this approach.
Deep learning models have reached superior results in various fields of application, but in many cases at a high cost of processing or large amount of data available. In most of them, specially in the medical field, the scarcity of training data limits the performance of these models. Among the strategies to overcome the lack of data, there is data augmentation, transfer learning and fine-tuning. In this work we compared different approaches to train deep convolutional neural network (CNN) to automatically detect architectural distortion (AD) in digital mammography. Although several computer vision based algorithms were designed to detect lesions in digital mammography, most of them perform poorly while detecting AD. We used the VGG-16 network pre-trained on ImageNet database with progressive fine-tuning to evaluate its performance on AD detection over a database of 280 images of clinical mammograms. Finally, we compared the results with a custom CNN architecture trained from scratch for the same task. Results indicated that a network with transfer learning and certain level of fine-tuning reaches the best results for the task (AUC = 0.89) compared with the other approaches, but no statistically significant difference was found between the best results using different amount of data augmentation and also compared to the custom CNN.
In previous work, we investigated the application of the normalized anisotropic quality index (NAQI) as an image quality metric for digital mammography. The initial assessment showed that NAQI depends not only on radiation dose, but also varies based on image features such as breast anatomy. In this work, these dependencies are analyzed by assessing the contribution of a range of features on NAQI values. The generalized matrix learning vector quantization (GMLVQ) was used to evaluate feature relevance and to rank the imaging parameters and breast features that affect NAQI. The GMLVQ uses prototype vectors to segregate and to analyze the NAQI in three classes: (1) low, (2) medium, and (3) high NAQI values. We used Spearman’s correlation coefficient (ρ) to compare the results obtained by the GMLVQ method. The GMLVQ was trained using 6,076 clinical mammograms. The statistical analysis showed that NAQI is dependent on several imaging parameters and breast features; in particular, breast area (ρ = -0.65), breast density (p = 0.62) and tube current-exposure time product (mAs) (p = 0.56). The GMLVQ results show that the most relevant parameters that affect the NAQI values were breast area (approx. 31%), mAs (approx. 24%) and breast density (approx. 15%). The GMLVQ method allowed us to better understand the NAQI results and provide support for the use of this metric for image quality assessment in digital mammography.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.