scholarly journals Optimizing risk-based breast cancer screening policies with reinforcement learning

2021 ◽  
Author(s):  
Adam Yala ◽  
Peter Mikhael ◽  
Constance Lehman ◽  
Gigin Lin ◽  
Fredrik Strand ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Reinforcement Learning ◽  
Early Detection ◽  
Massachusetts General Hospital ◽  
Chang Gung Memorial Hospital ◽  
Risk Models ◽  
Patient Risk ◽  
Screening Programs ◽  
Screening Cost ◽  
Wide Range

Abstract Screening programs must balance the benefits of early detection against the costs of over screening. Achieving this goal relies on two complementary technologies: (1) the ability to assess patient risk, (2) the ability to develop personalized screening programs given that risk. While methodologies for assessing patient risk have significantly improved with new advances in deep learning applied to imaging and genetics, our ability to personalize screening policies still lags behind. Here, we introduce a novel reinforcement learning-based framework for personalized screening, Tempo, and demonstrate its efficacy in the context of breast cancer. We trained our risk-based screening policies on a large screening mammography dataset from Massachusetts General Hospital (MGH) USA and validated them on held-out patients from MGH, and on external datasets from Emory USA, Karolinska Sweden and Chang Gung Memorial Hospital (CGMH) Taiwan. Across all test sets, we found that a Tempo policy combined with an image-based AI risk model was significantly more efficient than current regimes used in clinical practice in terms of simulated early detection per screen frequency. Moreover, we showed that the same Tempo policy can be easily adapted to a wide range of possible screening preferences, allowing clinicians to select their desired early detection to screening cost trade-off without training a new policy. Finally, we demonstrated Tempo policies based on AI-based risk models out performed Tempo policies based on less accurate clinical risk models. Altogether, our results show that pairing AI-based risk models with agile AI-designed screening policies has the potential to improve screening programs, advancing early detection while reducing over-screening.

Multi-Institutional Validation of a Mammography-Based Breast Cancer Risk Model

2021 ◽  
Author(s):  
Adam Yala ◽  
Peter G. Mikhael ◽  
Fredrik Strand ◽  
Gigin Lin ◽  
Siddharth Satuluru ◽  
...  
Keyword(s):  
Breast Cancer ◽  
General Hospital ◽  
Risk Model ◽  
Massachusetts General Hospital ◽  
High Sensitivity ◽  
Memorial Hospital ◽  
Significant Advance ◽  
Chang Gung Memorial Hospital ◽  
Risk Models ◽  
Screening Programs

PURPOSE Accurate risk assessment is essential for the success of population screening programs in breast cancer. Models with high sensitivity and specificity would enable programs to target more elaborate screening efforts to high-risk populations, while minimizing overtreatment for the rest. Artificial intelligence (AI)-based risk models have demonstrated a significant advance over risk models used today in clinical practice. However, the responsible deployment of novel AI requires careful validation across diverse populations. To this end, we validate our AI-based model, Mirai, across globally diverse screening populations. METHODS We collected screening mammograms and pathology-confirmed breast cancer outcomes from Massachusetts General Hospital, USA; Novant, USA; Emory, USA; Maccabi-Assuta, Israel; Karolinska, Sweden; Chang Gung Memorial Hospital, Taiwan; and Barretos, Brazil. We evaluated Uno's concordance-index for Mirai in predicting risk of breast cancer at one to five years from the mammogram. RESULTS A total of 128,793 mammograms from 62,185 patients were collected across the seven sites, of which 3,815 were followed by a cancer diagnosis within 5 years. Mirai obtained concordance indices of 0.75 (95% CI, 0.72 to 0.78), 0.75 (95% CI, 0.70 to 0.80), 0.77 (95% CI, 0.75 to 0.79), 0.77 (95% CI, 0.73 to 0.81), 0.81 (95% CI, 0.79 to 0.82), 0.79 (95% CI, 0.76 to 0.83), and 0.84 (95% CI, 0.81 to 0.88) at Massachusetts General Hospital, Novant, Emory, Maccabi-Assuta, Karolinska, Chang Gung Memorial Hospital, and Barretos, respectively. CONCLUSION Mirai, a mammography-based risk model, maintained its accuracy across globally diverse test sets from seven hospitals across five countries. This is the broadest validation to date of an AI-based breast cancer model and suggests that the technology can offer broad and equitable improvements in care.

Toward robust mammography-based models for breast cancer risk

2021 ◽  
Vol 13 (578) ◽  
pp. eaba4373 ◽  
Author(s):  
Adam Yala ◽  
Peter G. Mikhael ◽  
Fredrik Strand ◽  
Gigin Lin ◽  
Kevin Smith ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Deep Learning ◽  
High Risk ◽  
Breast Cancer Risk ◽  
Cancer Risk ◽  
Massachusetts General Hospital ◽  
The United States ◽  
University Hospital ◽  
Chang Gung Memorial Hospital ◽  
Risk Models

Improved breast cancer risk models enable targeted screening strategies that achieve earlier detection and less screening harm than existing guidelines. To bring deep learning risk models to clinical practice, we need to further refine their accuracy, validate them across diverse populations, and demonstrate their potential to improve clinical workflows. We developed Mirai, a mammography-based deep learning model designed to predict risk at multiple timepoints, leverage potentially missing risk factor information, and produce predictions that are consistent across mammography machines. Mirai was trained on a large dataset from Massachusetts General Hospital (MGH) in the United States and tested on held-out test sets from MGH, Karolinska University Hospital in Sweden, and Chang Gung Memorial Hospital (CGMH) in Taiwan, obtaining C-indices of 0.76 (95% confidence interval, 0.74 to 0.80), 0.81 (0.79 to 0.82), and 0.79 (0.79 to 0.83), respectively. Mirai obtained significantly higher 5-year ROC AUCs than the Tyrer-Cuzick model (P < 0.001) and prior deep learning models Hybrid DL (P < 0.001) and Image-Only DL (P < 0.001), trained on the same dataset. Mirai more accurately identified high-risk patients than prior methods across all datasets. On the MGH test set, 41.5% (34.4 to 48.5) of patients who would develop cancer within 5 years were identified as high risk, compared with 36.1% (29.1 to 42.9) by Hybrid DL (P = 0.02) and 22.9% (15.9 to 29.6) by the Tyrer-Cuzick model (P < 0.001).

scholarly journals Attributes, Performance, and Gaps in Current & Emerging Breast Cancer Screening Technologies

2019 ◽  
Vol 15 (2) ◽  
pp. 122-131
Author(s):  
Hina M. Ismail ◽  
Christopher G. Pretty ◽  
Matthew K. Signal ◽  
Marcus Haggers ◽  
J. Geoffrey Chase
Keyword(s):  
Breast Cancer ◽  
Cancer Screening ◽  
Early Detection ◽  
Breast Cancer Screening ◽  
Screening Tests ◽  
Screening Methods ◽  
Extensive Literature ◽  
Screening Programs ◽  
Detection Techniques ◽  
Wide Range

Background:Early detection of breast cancer, combined with effective treatment, can reduce mortality. Millions of women are diagnosed with breast cancer and many die every year globally. Numerous early detection screening tests have been employed. A wide range of current breast cancer screening methods are reviewed based on a series of searchers focused on clinical testing and performance. </P><P> Discussion: The key factors evaluated centre around the trade-offs between accuracy (sensitivity and specificity), operator dependence of results, invasiveness, comfort, time required, and cost. All of these factors affect the quality of the screen, access/eligibility, and/or compliance to screening programs by eligible women. This survey article provides an overview of the working principles, benefits, limitations, performance, and cost of current breast cancer detection techniques. It is based on an extensive literature review focusing on published works reporting the main performance, cost, and comfort/compliance metrics considered.Conclusion:Due to limitations and drawbacks of existing breast cancer screening methods there is a need for better screening methods. Emerging, non-invasive methods offer promise to mitigate the issues particularly around comfort/pain and radiation dose, which would improve compliance and enable all ages to be screened regularly. However, these methods must still undergo significant validation testing to prove they can provide realistic screening alternatives to the current accepted standards.

scholarly journals A limited review of over diagnosis methods and long-term effects in breast cancer screening

2011 ◽  
pp. 143-147
Author(s):  
Dongfeng Wu ◽  
Adriana Pérez
Keyword(s):  
Breast Cancer ◽  
Cancer Screening ◽  
Early Detection ◽  
Breast Cancer Screening ◽  
Screening Program ◽  
Critical Issue ◽  
Long Term Effects ◽  
Screening Programs ◽  
Probability Modeling

Breast cancer screening programs have been effective in detecting tumors prior to symptoms. Recently, there has been concern over the issue of over-diagnosis, that is, diagnosis of a breast cancer that does not manifest prior to death. Estimates for over-diagnosis vary, ranging from 7 to 52%. This variability may be due partially to issues associated with bias and/or incorrect inferences associated with the lack of probability modeling. A critical issue is how to evaluate the long-term effects due to continued screening. Participants in a periodic screening program can be classified into four mutually exclusive groups depending on whether individuals are diagnosed and whether their symptoms appear prior to death: True-earlydetection; No-early-detection; Over-diagnosis; and Not-sonecessary. All initially superficially healthy people will eventually fall into one of these four categories. This manuscript reviews the major methodologies associated with the over-diagnosis and long-term effects of breast cancer screening.

scholarly journals Increase Coverage Cancer Registry by Strengthening and Improving the Reporting of Breast Cancer Screening Program in Jakarta as One of Data Source

2018 ◽  
Vol 4 (Supplement 2) ◽  
pp. 142s-142s ◽  
Author(s):  
D.T. Sinulingga ◽  
A. Kadir ◽  
D. Purwanto ◽  
K. Kardinah ◽  
E. Suzanna ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Early Detection ◽  
Cancer Registry ◽  
Registry Data ◽  
Cancer Registration ◽  
Birth Date ◽  
Screening Programs ◽  
Cancer Registry Data ◽  
Mobile Mammography ◽  
Data Source

Background and context: Breast cancer screening programs by mobile mammography have been done since 2005 for Jakarta and around by collaboration of Dharmais National Cancer Center with The Indonesia Breast Cancer Foundation and Jakarta Health Department. Every activity, mobile mammography has been examined 50 persons for 60-80 times yearly. For 2015 there were 3493 examinations with 529 cases (15.1%) were abnormal which 43 from that cases (8.1%) are suspect malignancy. All that data were compile with early detection and cancer registry data to know the real cancer, but there were no malignancy and only 15 benign cases had come to follow-up to Dharmais NCC. The result is no malignancy. To compile the mobile mammography data with detection unit and cancer registry data were difficult because of the difference terms of the variables. National Cancer Registry of Indonesia has been announced by Ministry of Health since 2016. The coverage of cancer that diagnosed in 2008-2012 in Jakarta just 30.7%. One of the sources of data are screening data. In Indonesia, screening programs for breast and cervix cancer have been done sporadically, including at Jakarta. Unfortunately for 2008-2012 diagnose year, there were no data can compile to cancer registry because there were no address data and name. So we collaborate to improve the data variables for cancer registration system. Aim: To conduct the mobile mammography screening data as one of the sources data for cancer registration to improve the coverage in Jakarta and to prepare the breast specific cancer registry system. Strategy/Tactics: Cancer Registry, Early Detection, Radiology and Pathology Unit was collaborating with The Indonesia Breast Cancer Foundation to improve the variables. Program/Policy process: We to improve the variables and their operational definition especially for name, address and birth date include risk factors and physical examinations variables. The palpations of breast was conduct by midwives or doctors before the examination with mammography. Cancer registration variables definition is the standard, so the name variable is the name that written in the National Identity Card, include the birth date and address. Information system department in Dharmais NCC has made the program to this data system and have been take place since July 2017. Outcomes: There were 1462 data that have been in hospital data based for mobile mammography with 237 cases (16.2%) were abnormal which 14 cases (5.9%) are suspect malignancy. Only 5 suspect malignancy cases had more examination in Dharmais NCC and all of them were diagnosed breast malignancy. What was learned: Standardization of variable definitions is very important for cancer registry data source to improve the coverage especially for early stage finding cases. But to know the standard diagnose and to follow the cases real conditions, we have to make a good and clear referral system networking.

scholarly journals Breast Cancer Beliefs as Potential Targets for Breast Cancer Awareness Efforts to Decrease Late-Stage Presentation in Uganda

2018 ◽  
pp. 1-9 ◽  
Author(s):  
John R. Scheel ◽  
Yamile Molina ◽  
Benjamin O. Anderson ◽  
Donald L. Patrick ◽  
Gertrude Nakigudde ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Early Detection ◽  
Community Sample ◽  
Demographic Factors ◽  
Explanatory Models ◽  
Cancer Knowledge ◽  
Cancer Awareness ◽  
Screening Programs ◽  
Breast Cancer Awareness ◽  
Cancer Beliefs

Purpose To assess breast cancer beliefs in Uganda and determine whether these beliefs are associated with factors potentially related to nonparticipation in early detection. Methods A survey with open- and close-ended items was conducted in a community sample of Ugandan women to assess their beliefs about breast cancer. Linear regression was used to ascertain associations between breast cancer beliefs and demographic factors potentially associated with early detection, including socioeconomic factors, health care access, prior breast cancer knowledge, and personal detection practices. Results Of the 401 Ugandan women surveyed, most had less than a primary school education and received medical care at community health centers. Most women either believed in or were unsure about cultural explanatory models for developing breast cancer (> 82%), and the majority listed these beliefs as the most important causes of breast cancer (69%). By comparison, ≤ 45% of women believed in scientific explanatory risks for developing breast cancer. Although most believed that regular screening and early detection would find breast cancer when it is easy to treat (88% and 80%, respectively), they simultaneously held fatalistic attitudes toward their own detection efforts, including belief or uncertainty that a cure is impossible once they could self-detect a lump (54%). Individual beliefs were largely independent of demographic factors. Conclusion Misconceptions about breast cancer risks and benefits of early detection are widespread in Uganda and must be addressed in future breast cancer awareness efforts. Until screening programs exist, most breast cancer will be self-detected. Unless addressed by future awareness efforts, the high frequency of fatalistic attitudes held by women toward their own detection efforts will continue to be deleterious to breast cancer early detection in sub-Saharan countries like Uganda.

scholarly journals A limited review of over diagnosis methods and long-term effects in breast cancer screening

2011 ◽  
Vol 5 (3) ◽  
pp. 143 ◽  
Author(s):  
Dongfeng Wu ◽  
Adriana Pérez
Keyword(s):  
Breast Cancer ◽  
Cancer Screening ◽  
Early Detection ◽  
Breast Cancer Screening ◽  
Screening Program ◽  
Critical Issue ◽  
Long Term Effects ◽  
Screening Programs ◽  
Probability Modeling

Breast cancer screening programs have been effective in detecting tumors prior to symptoms. Recently, there has been concern over the issue of over-diagnosis, that is, diagnosis of a breast cancer that does not manifest prior to death. Estimates for over-diagnosis vary, ranging from 7 to 52%. This variability may be due partially to issues associated with bias and/or incorrect inferences associated with the lack of probability modeling. A critical issue is how to evaluate the long-term effects due to continued screening. Participants in a periodic screening program can be classified into four mutually exclusive groups depending on whether individuals are diagnosed and whether their symptoms appear prior to death: True-earlydetection; No-early-detection; Over-diagnosis; and Not-sonecessary. All initially superficially healthy people will eventually fall into one of these four categories. This manuscript reviews the major methodologies associated with the over-diagnosis and long-term effects of breast cancer screening.

scholarly journals Breast Cancer Screening Among Medical Staff in Tehran, Iran: Improvement of Early Detection

2020 ◽  
Author(s):  
Hossein Yahyazadeh ◽  
Marzieh Beheshti ◽  
Azita Abdollahinejad ◽  
Maria Hashemian ◽  
Narges Sistany Allahabad ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Cancer Screening ◽  
Early Detection ◽  
Breast Cancer Screening ◽  
Medical Staff ◽  
Demographic Data ◽  
Cross Sectional Study ◽  
Breast Diseases ◽  
Cross Sectional ◽  
Screening Programs

Breast cancer is the most prevalent Iranian female malignancy. Breast screening reduces the number of malignant breast diseases. We aimed to assess the results of the pilot breast cancer screening on early detection in female medical staff in Milad Hospital, Tehran, Iran. A cross-sectional study. Female medical staff from Milad Hospital, Tehran, Iran, were examined by a specialist in 2016. A checklist, including demographic data, was completed by the participants. If necessary, they referred to as sonography or mammography. Data were analyzed using SPSS software. Of 746 people enrolled, 137 had no pathological point, 609 had suspicious or positive findings that were referred for further investigation, 449 had normal findings, and 7 had suspicious mass and were biopsied, 6 were benign. One case had primary invasive cancer. Since screening for breast cancer helps to early detection of this disease, the implementation of cancer screening programs should be on the priority of health authorities.

scholarly journals Early Detection of Breast Cancer in Gulf Cooperation Council Countries: Case Studies

2018 ◽  
Vol 4 (Supplement 2) ◽  
pp. 50s-50s
Author(s):  
I. Fadhil ◽  
M. Alkuwari ◽  
F. Al Tahan ◽  
K. Alsaleh ◽  
D. Alsaadoon
Keyword(s):  
Breast Cancer ◽  
Early Detection ◽  
Case Studies ◽  
Professional Education ◽  
Breast Cancer Incidence ◽  
Advanced Stage ◽  
Gulf Region ◽  
Control Plan ◽  
Screening Programs ◽  
Gulf Countries

Background: Breast cancer is the most frequently diagnosed cancer among women in Gulf countries. Although breast cancer incidence rates in the Gulf are substantially lower than rates in developed countries, yet an increasing trend is evident. Most breast cancers are diagnosed at an advanced stage, only 23.3% of patients presented with localized tumors and less than 2% with in situ, making improvements to early detection of breast cancer a priority. There has been good progress and investment in early detection of breast cancer program in Gulf countries based on augmenting breast cancer awareness through public education, investing in mammographic based screening and improving infrastructure. Nevertheless, development of breast cancer early detection programs in most of the Gulf countries has been based on sporadic investments and actions rather than on a planned, approved and resource-linked national control plan. In many instances the scientific evidence-base for such investments has not been obtained and the evaluation of implemented programs is lacking. Aim: To review breast cancer screening, early detection practices in Gulf region, outlines enablers and identifies priorities for scaling up early detection programs in Gulf countries. Methods: The study relies heavily on review of published literature and data gathered through interview and discussion with key informants from government and nongovernment institutions at the studied countries. Results: Four case studies will be discussed from Bahrain, UAE, Kuwait and Saudi Arabia. Conclusion: Breast cancer is a major and increasing problem in Gulf countries, but it is still largely diagnosed at an advanced stage. While mammography based screening programs have been initiated in Gulf countries, however they generally have limited uptake, with very little evidence to support their effectiveness, largely because their attempts at education on the curability of breast cancer, and their endeavors to dispel the prevalent myths on breast cancer, have not been sufficiently successful. Thus, it is essential that the highest priority in each country should be improving awareness, early diagnosis of breast cancer, by public and professional education. This will require considerable investment in training primary care professionals, organizing referral mechanisms and setting up multidisciplinary breast cancer diagnosis and treatment facilities across the countries. While population-level screening for breast cancer is feasible in Gulf countries, yet careful consideration for available resources is critical for success. Moreover, it is important to pilot any screening programs prior to national roll-out.

scholarly journals The efficacy of using computer-aided detection (CAD) for detection of breast cancer in mammography screening: a systematic review

2018 ◽  
Vol 60 (1) ◽  
pp. 13-18 ◽  
Author(s):  
Emilie L Henriksen ◽  
Jonathan F Carlsen ◽  
Ilse MM Vejborg ◽  
Michael B Nielsen ◽  
Carsten A Lauridsen
Keyword(s):  
Breast Cancer ◽  
Early Detection ◽  
High Volume ◽  
Population Based ◽  
Screening Mammography ◽  
Computer Aided Detection ◽  
Screening Programs ◽  
Significant Difference ◽  
Computer Aided ◽  
Screening Population

Background Early detection of breast cancer (BC) is crucial in lowering the mortality. Purpose To present an overview of studies concerning computer-aided detection (CAD) in screening mammography for early detection of BC and compare diagnostic accuracy and recall rates (RR) of single reading (SR) with SR + CAD and double reading (DR) with SR + CAD. Material and Methods PRISMA guidelines were used as a review protocol. Articles on clinical trials concerning CAD for detection of BC in a screening population were included. The literature search resulted in 1522 records. A total of 1491 records were excluded by abstract and 18 were excluded by full text reading. A total of 13 articles were included. Results All but two studies from the SR vs. SR + CAD group showed an increased sensitivity and/or cancer detection rate (CDR) when adding CAD. The DR vs. SR + CAD group showed no significant differences in sensitivity and CDR. Adding CAD to SR increased the RR and decreased the specificity in all but one study. For the DR vs. SR + CAD group only one study reported a significant difference in RR. Conclusion All but two studies showed an increase in RR, sensitivity and CDR when adding CAD to SR. Compared to DR no statistically significant differences in sensitivity or CDR were reported. Additional studies based on organized population-based screening programs, with longer follow-up time, high-volume readers, and digital mammography are needed to evaluate the efficacy of CAD.

Sign in / Sign up

Export Citation Format

Share Document