Deep Convolutional Neural Network With Adversarial Training for Denoising Digital Breast Tomosynthesis Images

X-ray source motion blur modeling and deblurring with generative diffusion for digital breast tomosynthesis

Objective. Digital breast tomosynthesis (DBT) has significantly improved the diagnosis of breast cancer due to its high sensitivity and specificity in detecting breast lesions compared to two-dimensional mammography. However, one of the primary challenges in DBT is the image blur resulting from x-ray source motion, particularly in DBT systems with a source in continuous-motion mode. This motion-induced blur can degrade the spatial resolution of DBT images, potentially affecting the visibility of subtle lesions such as microcalcifications.Approach. We addressed this issue by deriving an analytical in-plane source blur kernel for DBT images based on imaging geometry and proposing a post-processing image deblurring method with a generative diffusion model as an image prior.Main results. We showed that the source blur could be approximated by a shift-invariant kernel over the DBT slice at a given height above the detector, and we validated the accuracy of our blur kernel modeling through simulation. We also demonstrated the ability of the diffusion model to generate realistic DBT images. The proposed deblurring method successfully enhanced spatial resolution when applied to DBT images reconstructed with detector blur and correlated noise modeling.Significance. Our study demonstrated the advantages of modeling the imaging system components such as source motion blur for improving DBT image quality.

Model-based deep CNN-regularized reconstruction for digital breast tomosynthesis with a task-based CNN image assessment approach

Objective. Digital breast tomosynthesis (DBT) is a quasi-three-dimensional breast imaging modality that improves breast cancer screening and diagnosis because it reduces fibroglandular tissue overlap compared with 2D mammography. However, DBT suffers from noise and blur problems that can lower the detectability of subtle signs of cancers such as microcalcifications (MCs). Our goal is to improve the image quality of DBT in terms of image noise and MC conspicuity.Approach. We proposed a model-based deep convolutional neural network (deep CNN or DCNN) regularized reconstruction (MDR) for DBT. It combined a model-based iterative reconstruction (MBIR) method that models the detector blur and correlated noise of the DBT system and the learning-based DCNN denoiser using the regularization-by-denoising framework. To facilitate the task-based image quality assessment, we also proposed two DCNN tools for image evaluation: a noise estimator (CNN-NE) trained to estimate the root-mean-square (RMS) noise of the images, and an MC classifier (CNN-MC) as a DCNN model observer to evaluate the detectability of clustered MCs in human subject DBTs.Main results. We demonstrated the efficacies of CNN-NE and CNN-MC on a set of physical phantom DBTs. The MDR method achieved low RMS noise and the highest detection area under the receiver operating characteristic curve (AUC) rankings evaluated by CNN-NE and CNN-MC among the reconstruction methods studied on an independent test set of human subject DBTs.Significance. The CNN-NE and CNN-MC may serve as a cost-effective surrogate for human observers to provide task-specific metrics for image quality comparisons. The proposed reconstruction method shows the promise of combining physics-based MBIR and learning-based DCNNs for DBT image reconstruction, which may potentially lead to lower dose and higher sensitivity and specificity for MC detection in breast cancer screening and diagnosis.

Recent advancements in artificial intelligence for breast cancer: Image augmentation, segmentation, diagnosis, and prognosis approaches

Breast cancer is a significant global health burden, with increasing morbidity and mortality worldwide. Early screening and accurate diagnosis are crucial for improving prognosis. Radiographic imaging modalities such as digital mammography (DM), digital breast tomosynthesis (DBT), magnetic resonance imaging (MRI), ultrasound (US), and nuclear medicine techniques, are commonly used for breast cancer assessment. And histopathology (HP) serves as the gold standard for confirming malignancy. Artificial intelligence (AI) technologies show great potential for quantitative representation of medical images to effectively assist in segmentation, diagnosis, and prognosis of breast cancer. In this review, we overview the recent advancements of AI technologies for breast cancer, including 1) improving image quality by data augmentation, 2) fast detection and segmentation of breast lesions and diagnosis of malignancy, 3) biological characterization of the cancer such as staging and subtyping by AI-based classification technologies, 4) prediction of clinical outcomes such as metastasis, treatment response, and survival by integrating multi-omics data. Then, we then summarize large-scale databases available to help train robust, generalizable, and reproducible deep learning models. Furthermore, we conclude the challenges faced by AI in real-world applications, including data curating, model interpretability, and practice regulations. Besides, we expect that clinical implementation of AI will provide important guidance for the patient-tailored management.

Deep learning denoising of digital breast tomosynthesis: Observer performance study of the effect on detection of microcalcifications in breast phantom images

BACKGROUND

The noise in digital breast tomosynthesis (DBT) includes x-ray quantum noise and detector readout noise. The total radiation dose of a DBT scan is kept at about the level of a digital mammogram but the detector noise is increased due to acquisition of multiple projections. The high noise can degrade the detectability of subtle lesions, specifically microcalcifications (MCs).

PURPOSE

We previously developed a deep-learning-based denoiser to improve the image quality of DBT. In the current study, we conducted an observer performance study with breast radiologists to investigate the feasibility of using deep-learning-based denoising to improve the detection of MCs in DBT.

METHODS

We have a modular breast phantom set containing seven 1-cm-thick heterogeneous 50% adipose/50% fibroglandular slabs custom-made by CIRS, Inc. (Norfolk, VA). We made six 5-cm-thick breast phantoms embedded with 144 simulated MC clusters of four nominal speck sizes (0.125-0.150, 0.150-0.180, 0.180-0.212, 0.212-0.250 mm) at random locations. The phantoms were imaged with a GE Pristina DBT system using the automatic standard (STD) mode. The phantoms were also imaged with the STD+ mode that increased the average glandular dose by 54% to be used as a reference condition for comparison of radiologists' reading. Our previously trained and validated denoiser was deployed to the STD images to obtain a denoised DBT set (dnSTD). Seven breast radiologists participated as readers to detect the MCs in the DBT volumes of the six phantoms under the three conditions (STD, STD+, dnSTD), totaling 18 DBT volumes. Each radiologist read all the 18 DBT volumes sequentially, which were arranged in a different order for each reader in a counter-balanced manner to minimize any potential reading order effects. They marked the location of each detected MC cluster and provided a conspicuity rating and their confidence level for the perceived cluster. The visual grading characteristics (VGC) analysis was used to compare the conspicuity ratings and the confidence levels of the radiologists for the detection of MCs.

RESULTS

The average sensitivities over all MC speck sizes were 65.3%, 73.2%, and 72.3%, respectively, for the radiologists reading the STD, dnSTD, and STD+ volumes. The sensitivity for dnSTD was significantly higher than that for STD (p < 0.005, two-tailed Wilcoxon signed rank test) and comparable to that for STD+. The average false positive rates were 3.9 ± 4.6, 2.8 ± 3.7, and 2.7 ± 3.9 marks per DBT volume, respectively, for reading the STD, dnSTD, and STD+ images but the difference between dnSTD and STD or STD+ did not reach statistical significance. The overall conspicuity ratings and confidence levels by VGC analysis for dnSTD were significantly higher than those for both STD and STD+ (p ≤ 0.001). The critical alpha value for significance was adjusted to be 0.025 with Bonferroni correction.

CONCLUSIONS

This observer study using breast phantom images showed that deep-learning-based denoising has the potential to improve the detection of MCs in noisy DBT images and increase radiologists' confidence in differentiating noise from MCs without increasing radiation dose. Further studies are needed to evaluate the generalizability of these results to the wide range of DBTs from human subjects and patient populations in clinical settings.

Task-based assessment of digital mammography microcalcification detection with deep learning denoising algorithmss using in silico and physical phantom studies

Purpose

Recent research suggests that image quality degradation with reduced radiation exposure in mammography can be mitigated by postprocessing mammograms with denoising algorithms based on convolutional neural networks. Breast microcalcifications, along with extended soft-tissue lesions, are the primary breast cancer biomarkers in a clinical x-ray examination, with the former being more sensitive to quantum noise. We test one such publicly available denoising method to observe if an improvement in detection of small microcalcifications can be achieved when deep learning-based denoising is applied to half-dose phantom scans.

Approach

An existing denoiser model (that was previously trained on clinical data) was applied to mammograms of an anthropomorphic physical phantom with hydroxyapatite microcalcifications. In addition, another model trained and tested using all synthetic (Monte Carlo) data was applied to a similar digital compressed breast phantom. Human reader studies were conducted to assess and compare image quality in a set of binary signal detection 4-AFC experiments, with proportion of correct responses used as a performance metric.

Results

In both physical phantom/clinical system and simulation studies, we saw no apparent improvement in small microcalcification signal detection in denoised half-dose mammograms. However, in a Monte Carlo study, we observed a noticeable jump in 4-AFC scores, when readers analyzed denoised half-dose images processed by the neural network trained on a dataset composed of 50% signal-present (SP) and 50% signal-absent regions of interest (ROIs).

Conclusions

Our findings conjecture that deep-learning denoising algorithms may benefit from enriching training datasets with SP ROIs, at least in cases with clusters of 5 to 10 microcalcifications, each of size ≲ 240    μ m .

Impact of loss functions on the performance of a deep neural network designed to restore low-dose digital mammography

Digital mammography is currently the most common imaging tool for breast cancer screening. Although the benefits of using digital mammography for cancer screening outweigh the risks associated with the x-ray exposure, the radiation dose must be kept as low as possible while maintaining the diagnostic utility of the generated images, thus minimizing patient risks. Many studies investigated the feasibility of dose reduction by restoring low-dose images using deep neural networks. In these cases, choosing the appropriate training database and loss function is crucial and impacts the quality of the results. In this work, we used a standard residual network (ResNet) to restore low-dose digital mammography images and evaluated the performance of several loss functions. For training purposes, we extracted 256,000 image patches from a dataset of 400 images of retrospective clinical mammography exams, where dose reduction factors of 75% and 50% were simulated to generate low and standard-dose pairs. We validated the network in a real scenario by using a physical anthropomorphic breast phantom to acquire real low-dose and standard full-dose images in a commercially available mammography system, which were then processed through our trained model. We benchmarked our results against an analytical restoration model for low-dose digital mammography. Objective assessment was performed through the signal-to-noise ratio (SNR) and the mean normalized squared error (MNSE), decomposed into residual noise and bias. Statistical tests revealed that the use of the perceptual loss (PL4) resulted in statistically significant differences when compared to all other loss functions. Additionally, images restored using the PL4 achieved the closest residual noise to the standard dose. On the other hand, perceptual loss PL3, structural similarity index (SSIM) and one of the adversarial losses achieved the lowest bias for both dose reduction factors. The source code of our deep neural network is available at https://github.com/WANG-AXIS/LdDMDenoising.

Radiation Dose Reduction in Digital Breast Tomosynthesis by MTANN with Multi-scale Kernels

Digital breast tomosynthesis (DBT) is an advanced three-dimensional screening modality for the early detection of breast cancer. DBT is able to reduce the problem of tissue overlap in standard two-dimensional mammograms, thus improving the sensitivity and specificity of cancer detection. Although DBT can improve diagnostic accuracy, it leads to higher radiation dose to patients compared to two-dimensional mammography. In this paper, we propose a novel radiation dose reduction technique that introduces multi-scale kernels to our original massive-training artificial neural network (MTANN) to reduce radiation dose substantially, while maintaining high image quality in DBT. After training our new MTANN with low-dose (LD) images and the corresponding "teaching" high-dose (HD) images, we can convert new LD images to "virtual" high-dose (VHD) images where noise and artifact in the LD images are significantly reduced. In VHD images, it is critical to preserve subtle structures and tiny patterns such as microcalcifications (MCs) which are essential for breast cancer diagnosis. We developed anatomical MTANN experts including an MC-specific expert with multi-scale kernels, which are combined by gating layers to generate whole VHD images. Our MTANN scheme was able to achieve a 79% dose reduction while preserving details of MCs. Experimental results demonstrated that our method achieved the highest performance among the best-known noise-reduction techniques and state-of-the-art deep-learning techniques.Clinical Relevance- Our method can decrease the dose radiation dose in DBT and maintain the image quality.

Denoising digital breast tomosynthesis projections using deep learning with synthetic data as training set

Purpose

Image denoising based on deep neural networks (DNN) needs a big dataset containing digital breast tomosynthesis (DBT) projections acquired in different radiation doses to be trained, which is impracticable. Therefore, we propose extensively investigating the use of synthetic data generated by software for training DNNs to denoise DBT real data.

Approach

The approach consists of generating a synthetic dataset representative of the DBT sample space by software, containing noisy and original images. Synthetic data were generated in two different ways: (a) virtual DBT projections generated by OpenVCT and (b) noisy images synthesized from photography regarding noise models used in DBT (e.g., Poisson-Gaussian noise). Then, DNN-based denoising techniques were trained using a synthetic dataset and tested for denoising physical DBT data. Results were evaluated in quantitative (PSNR and SSIM measures) and qualitative (visual analysis) terms. Furthermore, a dimensionality reduction technique (t-SNE) was used for visualization of sample spaces of synthetic and real datasets.

Results

The experiments showed that training DNN models with synthetic data could denoise DBT real data, achieving competitive results to traditional methods in quantitative terms but showing a better balance between noise filtering and detail preservation in a visual analysis. T-SNE enables us to visualize if synthetic and real noises are in the same sample space.

Conclusion

We propose a solution for the lack of suitable training data to train DNN models for denoising DBT projections, showing that we just need the synthesized noise to be in the same sample space as the target image.

Image Augmentation Techniques for Mammogram Analysis

Research in the medical imaging field using deep learning approaches has become progressively contingent. Scientific findings reveal that supervised deep learning methods’ performance heavily depends on training set size, which expert radiologists must manually annotate. The latter is quite a tiring and time-consuming task. Therefore, most of the freely accessible biomedical image datasets are small-sized. Furthermore, it is challenging to have big-sized medical image datasets due to privacy and legal issues. Consequently, not a small number of supervised deep learning models are prone to overfitting and cannot produce generalized output. One of the most popular methods to mitigate the issue above goes under the name of data augmentation. This technique helps increase training set size by utilizing various transformations and has been publicized to improve the model performance when tested on new data. This article surveyed different data augmentation techniques employed on mammogram images. The article aims to provide insights into basic and deep learning-based augmentation techniques.

Evaluation of a Generative Adversarial Network to Improve Image Quality and Reduce Radiation-Dose during Digital Breast Tomosynthesis

In this study, we evaluated the improvement of image quality in digital breast tomosynthesis under low-radiation dose conditions of pre-reconstruction processing using conditional generative adversarial networks [cGAN (pix2pix)]. Pix2pix pre-reconstruction processing with filtered back projection (FBP) was compared with and without multiscale bilateral filtering (MSBF) during pre-reconstruction processing. Noise reduction and preserve contrast rates were compared using full width at half-maximum (FWHM), contrast-to-noise ratio (CNR), peak signal-to-noise ratio (PSNR), and structural similarity (SSIM) in the in-focus plane using a BR3D phantom at various radiation doses [reference-dose (automatic exposure control reference dose: AECrd), 50% and 75% reduction of AECrd] and phantom thicknesses (40 mm, 50 mm, and 60 mm). The overall performance of pix2pix pre-reconstruction processing was effective in terms of FWHM, PSNR, and SSIM. At ~50% radiation-dose reduction, FWHM yielded good results independently of the microcalcification size used in the BR3D phantom, and good noise reduction and preserved contrast. PSNR results showed that pix2pix pre-reconstruction processing represented the minimum in the error with reference FBP images at an approximately 50% reduction in radiation-dose. SSIM analysis indicated that pix2pix pre-reconstruction processing yielded superior similarity when compared with and without MSBF pre-reconstruction processing at ~50% radiation-dose reduction, with features most similar to the reference FBP images. Thus, pix2pix pre-reconstruction processing is promising for reducing noise with preserve contrast and radiation-dose reduction in clinical practice.

Usefulness of a Metal Artifact Reduction Algorithm in Digital Tomosynthesis Using a Combination of Hybrid Generative Adversarial Networks

In this study, a novel combination of hybrid generative adversarial networks (GANs) comprising cycle-consistent GAN, pix2pix, and (mask pyramid network) MPN (CGpM-metal artifact reduction [MAR]), was developed using projection data to reduce metal artifacts and the radiation dose during digital tomosynthesis. The CGpM-MAR algorithm was compared with the conventional filtered back projection (FBP) without MAR, FBP with MAR, and convolutional neural network MAR. The MAR rates were compared using the artifact index (AI) and Gumbel distribution of the largest variation analysis using a prosthesis phantom at various radiation doses. The novel CGpM-MAR yielded an adequately effective overall performance in terms of AI. The resulting images yielded good results independently of the type of metal used in the prosthesis phantom (p < 0.05) and good artifact removal at 55% radiation-dose reduction. Furthermore, the CGpM-MAR represented the minimum in the model with the largest variation at 55% radiation-dose reduction. Regarding the AI and Gumbel distribution analysis, the novel CGpM-MAR yielded superior MAR when compared with the conventional reconstruction algorithms with and without MAR at 55% radiation-dose reduction and presented features most similar to the reference FBP. CGpM-MAR presents a promising method for metal artifact and radiation-dose reduction in clinical practice.