A Deep Learning Approach to Re-create Raw Full-Field Digital                     Mammograms for Breast Density and Texture Analysis

Shu, Hai; Chiang, Ting-Yu D.; Wei, Peng; Anh, Kim; Lesslie, Michele D.; Cohen, Ethan O.; Srinivasan, Ashmitha; Moseley, Tanya W.; Sen, Lauren Q. Chang; Leung, Jessica W.T.; Dennison, Jennifer B.; Hanash, Sam; Weaver, Olena

doi:10.1148/ryai.2021200097

Cited by 6 publications

(7 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…1000 of the 5032 participants have been previously reported using 4394 matched pairs of for processing and for presentation FFDM [ 21 ]. This prior study developed a DL-based approach to recreate for-processing FFDM from for-presentation mammograms.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning Models for Automated Assessment of Breast Density Using Multiple Mammographic Image Types

Rigaud

Weaver²,

Dennison³

et al. 2022

Cancers

View full text Add to dashboard Cite

Recently, convolutional neural network (CNN) models have been proposed to automate the assessment of breast density, breast cancer detection or risk stratification using single image modality. However, analysis of breast density using multiple mammographic types using clinical data has not been reported in the literature. In this study, we investigate pre-trained EfficientNetB0 deep learning (DL) models for automated assessment of breast density using multiple mammographic types with and without clinical information to improve reliability and versatility of reporting. 120,000 for-processing and for-presentation full-field digital mammograms (FFDM), digital breast tomosynthesis (DBT), and synthesized 2D images from 5032 women were retrospectively analyzed. Each participant underwent up to 3 screening examinations and completed a questionnaire at each screening encounter. Pre-trained EfficientNetB0 DL models with or without clinical history were optimized. The DL models were evaluated using BI-RADS (fatty, scattered fibroglandular densities, heterogeneously dense, or extremely dense) versus binary (non-dense or dense) density classification. Pre-trained EfficientNetB0 model performances were compared using inter-observer and commercial software (Volpara) variabilities. Results show that the average Fleiss’ Kappa score between-observers ranged from 0.31–0.50 and 0.55–0.69 for the BI-RADS and binary classifications, respectively, showing higher uncertainty among experts. Volpara-observer agreement was 0.33 and 0.54 for BI-RADS and binary classifications, respectively, showing fair to moderate agreement. However, our proposed pre-trained EfficientNetB0 DL models-observer agreement was 0.61–0.66 and 0.70–0.75 for BI-RADS and binary classifications, respectively, showing moderate to substantial agreement. Overall results show that the best breast density estimation was achieved using for-presentation FFDM and DBT images without added clinical information. Pre-trained EfficientNetB0 model can automatically assess breast density from any images modality type, with the best results obtained from for-presentation FFDM and DBT, which are the most common image archived in clinical practice.

show abstract

Section: Methodsmentioning

confidence: 99%

“…Each of the 11 models listed above was trained with both views (i.e., CC and MLO) and both sides (i.e., left and right). The background of the images was removed to avoid biased feature extraction outside the breast [ 21 ]. All image intensities were Z standardized and normalized between 0–255.…”

Section: Methodsmentioning

confidence: 99%

Deep Learning Models for Automated Assessment of Breast Density Using Multiple Mammographic Image Types

Rigaud

Weaver²,

Dennison³

et al. 2022

Cancers

View full text Add to dashboard Cite

show abstract

“…More recently, Lotter et al [ 6 ] applied the ResNet to over 100 000 2D and 3D screening mammograms from 5 sites in the US and China and the resulting deep learning model had AUCs ranging from 0.922 to 0.971 and outperformed five out of five radiologists, achieving a 14.2% greater sensitivity and a 24% increase in specificity. On the other hand, Shu et al [ 31 ] approached computational modeling of screening mammography from a different angle and developed a deep learning approach to re-create rarely stored for-processing (raw) digital mammograms from routinely stored for-presentation (processed) mammograms. ‘Raw’ images generated by full-field digital mammography are digitally manipulated to enhance some features, such as contrast and resolution, to produce ‘for-presentation’ images that are optimized for visual cancer detection by radiologists.…”

Section: Deep Learning and Its Applications In Medical Imagingmentioning

confidence: 99%

“…Moreover, mammography equipment manufacturers do not disclose their raw-to-processed-image conversion steps, and inversion algorithms are not available. To this end, Shu et al [ 31 ] developed a deep-learning approach, based on the U-Net CNN, to re-create raw digital mammograms from for-presentation mammograms. The authors used 3713 pairs of raw and processed mammograms collected from nearly 900 women in the ‘Mammography, Early Detection Biomarkers, Risk Assessment, and Imaging Technologies’ (MERIT) cohort study (ClinicalTrials.gov Identifier: NCT03408353) at the University of Texas MD Anderson Cancer Center.…”

Section: Deep Learning and Its Applications In Medical Imagingmentioning

confidence: 99%

“…Finally, deep learning is much more computationally expensive than traditional radiomics analysis [ 36 ]. Take, for example, the deep learning of screening mammograms [ 31 ]. While it took several hours to train the modified U-Net for 300 epochs (training cycles) using four NVIDIA Titan Xp GPUs (total memory, 48 GB) given a set of hyper-parameters, such as learning rate and mini-batch size, tuning the hyper-parameters entailed many rounds of re-training the model.…”

Section: Radiomics Vs Deep Learningmentioning

confidence: 99%

See 1 more Smart Citation

Radiomics, deep learning and early diagnosis in oncology

Wei

2021

Emerging Topics in Life Sciences

Self Cite

View full text Add to dashboard Cite

Medical imaging, including X-ray, computed tomography (CT), and magnetic resonance imaging (MRI), plays a critical role in early detection, diagnosis, and treatment response prediction of cancer. To ease radiologists’ task and help with challenging cases, computer-aided diagnosis has been developing rapidly in the past decade, pioneered by radiomics early on, and more recently, driven by deep learning. In this mini-review, I use breast cancer as an example and review how medical imaging and its quantitative modeling, including radiomics and deep learning, have improved the early detection and treatment response prediction of breast cancer. I also outline what radiomics and deep learning share in common and how they differ in terms of modeling procedure, sample size requirement, and computational implementation. Finally, I discuss the challenges and efforts entailed to integrate deep learning models and software in clinical practice.

show abstract

BiTr-Unet: A CNN-Transformer Combined Network for MRI Brain Tumor Segmentation

Qiran

Shu

2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

A Deep Learning Approach to Re-create Raw Full-Field Digital Mammograms for Breast Density and Texture Analysis

Cited by 6 publications

References 27 publications

Deep Learning Models for Automated Assessment of Breast Density Using Multiple Mammographic Image Types

Deep Learning Models for Automated Assessment of Breast Density Using Multiple Mammographic Image Types

Radiomics, deep learning and early diagnosis in oncology

BiTr-Unet: A CNN-Transformer Combined Network for MRI Brain Tumor Segmentation

Contact Info

Product

Resources

About