The Role of Deep Learning in Breast Screening

Harvey, Hugh; Karpati, Edith; Khara, Galvin; Korkinof, Dimitrios; Ng, Aik Seng; Austin, Christopher C.; Rijken, Tobias; Kecskemethy, Peter D.

doi:10.1007/s12609-019-0301-7

Cited by 23 publications

(14 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Some commentators are already speculating that this will lead to the dismantling of the traditional professions [ 30 ]. On the other hand, the prospect of the delegation of decision-making authority from humans to machines [ 31 ], or the outright substitution of fallible or scarce human expertise by more reliable AI tools [ 32 ], leaves others more sceptical. Some argue, for instance, that the context-sensitive intelligence and tacit knowledge of professionals still trumps the narrowly data-driven capabilities of AI systems [ 33 ].…”

Section: Discussionmentioning

confidence: 99%

Professionals’ responses to the introduction of AI innovations in radiology and their implications for future adoption: a qualitative study

Chen

Stavropoulou

Narasinkan

et al. 2021

BMC Health Serv Res

View full text Add to dashboard Cite

Background Artificial Intelligence (AI) innovations in radiology offer a potential solution to the increasing demand for imaging tests and the ongoing workforce crisis. Crucial to their adoption is the involvement of different professional groups, namely radiologists and radiographers, who work interdependently but whose perceptions and responses towards AI may differ. We aim to explore the knowledge, awareness and attitudes towards AI amongst professional groups in radiology, and to analyse the implications for the future adoption of these technologies into practice. Methods We conducted 18 semi-structured interviews with 12 radiologists and 6 radiographers from four breast units in National Health Services (NHS) organisations and one focus group with 8 radiographers from a fifth NHS breast unit, between 2018 and 2020. Results We found that radiographers and radiologists vary with respect to their awareness and knowledge around AI. Through their professional networks, conference attendance, and contacts with industry developers, radiologists receive more information and acquire more knowledge of the potential applications of AI. Radiographers instead rely more on localized personal networks for information. Our results also show that although both groups believe AI innovations offer a potential solution to workforce shortages, they differ significantly regarding the impact they believe it will have on their professional roles. Radiologists believe AI has the potential to take on more repetitive tasks and allow them to focus on more interesting and challenging work. They are less concerned that AI technology might constrain their professional role and autonomy. Radiographers showed greater concern about the potential impact that AI technology could have on their roles and skills development. They were less confident of their ability to respond positively to the potential risks and opportunities posed by AI technology. Conclusions In summary, our findings suggest that professional responses to AI are linked to existing work roles, but are also mediated by differences in knowledge and attitudes attributable to inter-professional differences in status and identity. These findings question broad-brush assertions about the future deskilling impact of AI which neglect the need for AI innovations in healthcare to be integrated into existing work processes subject to high levels of professional autonomy.

show abstract

Section: Discussionmentioning

confidence: 99%

Professionals’ responses to the introduction of AI innovations in radiology and their implications for future adoption: a qualitative study

Chen

Stavropoulou

Narasinkan

et al. 2021

BMC Health Serv Res

View full text Add to dashboard Cite

show abstract

“…As a result, our model has the ability to learn both local features across the entire image as well as macroscopic features such as symmetry between breasts. For a more comprehensive review of prior work, refer to one of the recent reviews [ 33 ], [ 34 ].…”

Section: Related Workmentioning

confidence: 99%

Deep Neural Networks Improve Radiologists’ Performance in Breast Cancer Screening

Phang

Park

et al. 2020

IEEE Trans. Med. Imaging

497

397

View full text Add to dashboard Cite

We present a deep convolutional neural network for breast cancer screening exam classification, trained and evaluated on over 200,000 exams (over 1,000,000 images). Our network achieves an AUC of 0.895 in predicting whether there is a cancer in the breast, when tested on the screening population. We attribute the high accuracy of our model to a two-stage training procedure, which allows us to use a very high-capacity patch-level network to learn from pixel-level labels alongside a network learning from macroscopic breast-level labels. To validate our model, we conducted a reader study with 14 readers, each reading 720 screening mammogram exams, and find our model to be as accurate as experienced radiologists when presented with the same data. Finally, we show that a hybrid model, averaging probability of malignancy predicted by a radiologist with a prediction of our neural network, is more accurate than either of the two separately. To better understand our results, we conduct a thorough analysis of our network's performance on different subpopulations of the screening population, model design, training procedure, errors, and properties of its internal representations.deep learning | deep convolutional neural networks | breast cancer screening | mammography B reast cancer is the second leading cancer-related cause of death among women in the US. In 2014, over 39 million screening and diagnostic mammography exams were performed in the US. It is estimated that in 2015 232,000 women were diagnosed with breast cancer and approximately 40,000 died from it (1). Although mammography is the only imaging test that has reduced breast cancer mortality (2-4), there has been discussion regarding the potential harms of screening, including false positive recalls and associated false positive biopsies. The vast majority of the 10-15% of women asked to return following an inconclusive screening mammogram undergo another mammogram and/or ultrasound for clarification. After the additional imaging exams, many of these findings are determined as benign and only 10-20% are recommended to undergo a needle biopsy for further work-up. Among these, only 20-40% yield a diagnosis of cancer (5). Evidently, there is an unmet need to shift the balance of routine breast cancer screening towards more benefit and less harm.Traditional computer-aided detection (CAD) in mammography is routinely used by radiologists to assist with image interpretation, despite multicenter studies showing these CAD programs do not improve their diagnostic performance (6).These CAD programs typically use handcrafted features to mark sites on a mammogram that appear distinct from normal tissue structures. The radiologist decides whether to recall these findings, determining clinical significance and actionability. Recent developments in deep learning (7)-in particular, deep convolutional neural networks (CNNs) (8-12)-open possibilities for creating a new generation of CAD-like tools.This paper makes several contributions. Primarily, we train and evaluate a set of stro...

show abstract

“…6 Internationally, there is increasing concern about the ongoing viability of population breast screening programmes due to what has been termed 'a global radiology workforce crisis'. 7 As in the UK and Europe, resourcing screen-reads in Australia is increasingly difficult for publicly funded screening programmes, where reader shortages exist in some locations. 8 The Royal Australian and New Zealand College of Radiologists' Workforce Survey Report identifies screening mammography as an area of practice 'at significant risk of workforce shortage', with this deficit predicted to increase over time.…”

Section: Strengths and Limitations Of This Studymentioning

confidence: 99%

Artificial intelligence (AI) to enhance breast cancer screening: protocol for population-based cohort study of cancer detection

Marinovich

Wylie

Lotter³

et al. 2022

BMJ Open

View full text Add to dashboard Cite

IntroductionArtiﬁcial intelligence (AI) algorithms for interpreting mammograms have the potential to improve the effectiveness of population breast cancer screening programmes if they can detect cancers, including interval cancers, without contributing substantially to overdiagnosis. Studies suggesting that AI has comparable or greater accuracy than radiologists commonly employ ‘enriched’ datasets in which cancer prevalence is higher than in population screening. Routine screening outcome metrics (cancer detection and recall rates) cannot be estimated from these datasets, and accuracy estimates may be subject to spectrum bias which limits generalisabilty to real-world screening. We aim to address these limitations by comparing the accuracy of AI and radiologists in a cohort of consecutive of women attending a real-world population breast cancer screening programme.Methods and analysisA retrospective, consecutive cohort of digital mammography screens from 109 000 distinct women was assembled from BreastScreen WA (BSWA), Western Australia’s biennial population screening programme, from November 2016 to December 2017. The cohort includes 761 screen-detected and 235 interval cancers. Descriptive characteristics and results of radiologist double-reading will be extracted from BSWA outcomes data collection. Mammograms will be reinterpreted by a commercial AI algorithm (DeepHealth). AI accuracy will be compared with that of radiologist single-reading based on the diﬀerence in the area under the receiver operating characteristic curve. Cancer detection and recall rates for combined AI–radiologist reading will be estimated by pairing the first radiologist read per screen with the AI algorithm, and compared with estimates for radiologist double-reading.Ethics and disseminationThis study has ethical approval from the Women and Newborn Health Service Ethics Committee (EC00350) and the Curtin University Human Research Ethics Committee (HRE2020-0316). Findings will be published in peer-reviewed journals and presented at national and international conferences. Results will also be disseminated to stakeholders in Australian breast cancer screening programmes and policy makers in population screening.

show abstract

The Role of Deep Learning in Breast Screening

Cited by 23 publications

References 31 publications

Professionals’ responses to the introduction of AI innovations in radiology and their implications for future adoption: a qualitative study

Professionals’ responses to the introduction of AI innovations in radiology and their implications for future adoption: a qualitative study

Deep Neural Networks Improve Radiologists’ Performance in Breast Cancer Screening

Artificial intelligence (AI) to enhance breast cancer screening: protocol for population-based cohort study of cancer detection

Contact Info

Product

Resources

About