Deep learning and three-compartment breast imaging in breast cancer diagnosis

Drukker, Karen; Huynh, Benjamin Q.; Giger, Maryellen L.; Malkov, Serghei; Avila, Jesus; Fan, Bo; Joe, Bonnie N.; Kerlikowske, Karla; Drukteinis, Jennifer S.; Kazemi, Leila; Pereira, Malesa; Shepherd, John

doi:10.1117/12.2254516

Cited by 4 publications

(5 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, ConvNets have found many applications in the field of medical imaging with an increasing number of publications. For example, ConvNets were used in cancer imaging for breast cancer [27, 28], brain tumor [29, 30] and lung nodule [31–33] detection, for classification of interstitial lung disease [34, 35] or for breast arterial calcifications [36]. In cardiac imaging, ConvNets have been widely used for ventricle and heart segmentation in MRI [37–40].…”

Section: Related Workmentioning

confidence: 99%

Deep Learning for Quantification of Epicardial and Thoracic Adipose Tissue From Non-Contrast CT

Commandeur

Goeller

Betancur

et al. 2018

IEEE Trans. Med. Imaging

152

100

View full text Add to dashboard Cite

Epicardial adipose tissue (EAT) is a visceral fat deposit related to coronary artery disease. Fully automated quantification of EAT volume in clinical routine could be a timesaving and reliable tool for cardiovascular risk assessment. We propose a new fully automated deep learning framework for EAT and thoracic adipose tissue (TAT) quantification from non-contrast coronary artery calcium computed tomography (CT) scans. The first multi-task convolutional neural network (ConvNet) is used to determine heart limits and perform segmentation of heart and adipose tissues. The second ConvNet, combined with a statistical shape model, allows for pericardium detection. EAT and TAT segmentations are then obtained from outputs of both ConvNets. We evaluate the performance of the method on CT data sets from 250 asymptomatic individuals. Strong agreement between automatic and expert manual quantification is obtained for both EAT and TAT with median Dice score coefficients of 0.823 (inter-quartile range (IQR): 0.779-0.860) and 0.905 (IQR: 0.862-0.928), respectively; with excellent correlations of 0.924 and 0.945 for EAT and TAT volumes. Computations are performed in <6 s on a standard personal computer for one CT scan. Therefore, the proposed method represents a tool for rapid fully automated quantification of adipose tissue and may improve cardiovascular risk stratification in patients referred for routine CT calcium scans.

show abstract

Section: Related Workmentioning

confidence: 99%

Deep Learning for Quantification of Epicardial and Thoracic Adipose Tissue From Non-Contrast CT

Commandeur

Goeller

Betancur

et al. 2018

IEEE Trans. Med. Imaging

152

100

View full text Add to dashboard Cite

show abstract

“…Beyond full-field digital mammograms and breast ultrasound, CAD on breast tomosynthesis was an early topic for emerging technology. [52][53][54] Other breast imaging technologies and topics with CAD applications included dynamic breast MRI, utilization of multiple views, lesion segmentation and classification, breast segmentation and density [55][56][57][58][59][60][61][62] In addition, AI methods for assessing prognosis and response to therapy have been presented. 63 Abdominal imaging with a focus on bowel and liver was a frequent topic.…”

Section: Introductionmentioning

confidence: 99%

“… 52 – 54 Other breast imaging technologies and topics with CAD applications included dynamic breast MRI, utilization of multiple views, lesion segmentation and classification, breast segmentation and density assessment, predictive models for cancer risk assessment, dedicated breast CT, 3D ultrasound, and breast cancer diagnosis with deep learning. 55 – 62 In addition, AI methods for assessing prognosis and response to therapy have been presented. 63 …”

Section: Introductionmentioning

confidence: 99%

SPIE Computer-Aided Diagnosis conference anniversary review

Summers

Giger

2022

J. Med. Imag.

Self Cite

View full text Add to dashboard Cite

The SPIE Computer-Aided Diagnosis conference has been held for 16 consecutive years at the annual SPIE Medical Imaging symposium. The conference remains vibrant, with a core group of submitters as well as new submitters and attendees each year. Recent developments include a marked shift in submissions relating to the artificial intelligence revolution in medical image analysis. This review describes the topics and trends observed in research presented at the Computer-Aided Diagnosis conference as part of the 50th-anniversary celebration of SPIE Medical Imaging.

show abstract

“…For the case of breast cancer diagnosis, the Breast Imaging Reporting and Data System (BI-RADS) is defined by the American College of Radiology (ACR) and widely used as a standardized method to record and communicate the abnormalities. 1 Recently, a deep learning technology has dramatically succeed in various applications such as image/video recognition, 2-4 biometrics, [5][6][7][8] image generation, [9][10][11] and medical image analysis [12][13][14][15] as well. Deep learning approaches have achieved impressive accuracies in computer-aided detection (CADe) and computer-aided diagnosis (CADx) on various modalities.…”

Section: Introductionmentioning

confidence: 99%

“…Deep learning approaches have achieved impressive accuracies in computer-aided detection (CADe) and computer-aided diagnosis (CADx) on various modalities. 12,13,15 The use of the current deep learning approaches for CADx in real world is limited due to the lack of interpretability. So, the interpretability of the decisions made by the deep learning approach needs to be investigated for real world deployment.…”

Section: Introductionmentioning

confidence: 99%

ICADx: interpretable computer aided diagnosis of breast masses

Kim

Lee

Kim

et al. 2018

Medical Imaging 2018: Computer-Aided Diagnosis

View full text Add to dashboard Cite

In this study, a novel computer aided diagnosis (CADx) framework is devised to investigate interpretability for classifying breast masses. Recently, a deep learning technology has been successfully applied to medical image analysis including CADx. Existing deep learning based CADx approaches, however, have a limitation in explaining the diagnostic decision. In real clinical practice, clinical decisions could be made with reasonable explanation. So current deep learning approaches in CADx are limited in real world deployment. In this paper, we investigate interpretability in CADx with the proposed interpretable CADx (ICADx) framework. The proposed framework is devised with a generative adversarial network, which consists of interpretable diagnosis network and synthetic lesion generative network to learn the relationship between malignancy and a standardized description (BI-RADS). The lesion generative network and the interpretable diagnosis network compete in an adversarial learning so that the two networks are improved. The effectiveness of the proposed method was validated on public mammogram database. Experimental results showed that the proposed ICADx framework could provide the interpretability of mass as well as mass classification. It was mainly attributed to the fact that the proposed method was effectively trained to find the relationship between malignancy and interpretations via the adversarial learning. These results imply that the proposed ICADx framework could be a promising approach to develop the CADx system.

show abstract

Deep learning and three-compartment breast imaging in breast cancer diagnosis

Cited by 4 publications

References 14 publications

Deep Learning for Quantification of Epicardial and Thoracic Adipose Tissue From Non-Contrast CT

Deep Learning for Quantification of Epicardial and Thoracic Adipose Tissue From Non-Contrast CT

SPIE Computer-Aided Diagnosis conference anniversary review

ICADx: interpretable computer aided diagnosis of breast masses

Contact Info

Product

Resources

About