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

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.

Download Full-text

Optimizing risk-based breast cancer screening policies with reinforcement learning

10.21203/rs.3.rs-104592/v1 ◽

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.

Download Full-text

Toward robust mammography-based models for breast cancer risk

Science Translational Medicine ◽

10.1126/scitranslmed.aba4373 ◽

2021 ◽

Vol 13 (578) ◽

pp. eaba4373 ◽

Cited By ~ 1

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).

Download Full-text

Diagnostic NeuroradiologyDiagnostic Neuroradiology. By TaverasJuan M.M.D. Professor of Radiology, College of Physicians and Surgeons, Columbia University; Attending Radiologist, Neurological Institute, Presbyterian Hospital of New York, and WoodErnest H.M.D. Professor and Chairman, Department of Radiology, School of Medicine, University of North Carolina; Director of Radiological Service, North Carolina Memorial Hospital; formerly Attending Radiologist, Neurological Institute, Presbyterian Hospital of New York. Reprinted fromGolden's Diagnostic Roentgenology, edited by RobbinsLaurence L. M.D., Clinical Professor of Radiology, Harvard Medical School; Radiologist-in-Chief, Massachusetts General Hospital, Boston, Mass. A volume of pages, 960 with 722 figures. Published by The Williams & Wilkins Co., Baltimore, Md, 1964. Price $32.50.

Radiology ◽

10.1148/83.5.950a ◽

1964 ◽

Vol 83 (5) ◽

pp. 950-950

Keyword(s):

New York ◽

North Carolina ◽

Medical School ◽

General Hospital ◽

Harvard Medical School ◽

Massachusetts General Hospital ◽

Memorial Hospital ◽

Columbia University ◽

Presbyterian Hospital ◽

School Of Medicine

Download Full-text

Slipped Capital Femoral EpiphysisBy KleinArmin, M.D., JoplinRobert J., M.D., ReidyJohn A., M.D., and HanelinJoseph, M.D., Massachusetts General Hospital, Boston, Mass. A monograph in American Lectures in Roentgen Diagnosis, edited by HamptonAubrey O., M.D., Chief, Department of Radiology, Garfield Memorial Hospital, Washington, D. C. A volume of 130 pages, with 230 illustrations. Published by Charles C Thomas, Springfield, Ill., 1953. Price $6.75.

Radiology ◽

10.1148/61.5.820a ◽

1953 ◽

Vol 61 (5) ◽

pp. 820-820

Keyword(s):

General Hospital ◽

Massachusetts General Hospital ◽

Memorial Hospital ◽

Roentgen Diagnosis

Download Full-text

The Advent of Salivary Breast Cancer Biomarker Detection Using Affinity Sensors

Sensors ◽

10.3390/s19102373 ◽

2019 ◽

Vol 19 (10) ◽

pp. 2373 ◽

Cited By ~ 8

Author(s):

Imad Abrao Nemeir ◽

Joseph Saab ◽

Walid Hleihel ◽

Abdelhamid Errachid ◽

Nicole Jafferzic-Renault ◽

...

Keyword(s):

Breast Cancer ◽

Low Cost ◽

Current Trend ◽

High Sensitivity ◽

Breast Cancer Diagnosis ◽

Current Status ◽

Attractive Alternative ◽

Early Screening ◽

Screening Programs ◽

Sensitivity And Selectivity

Breast Cancer is one of the world’s most notorious diseases affecting two million women in 2018 worldwide. It is a highly heterogeneous disease, making it difficult to treat. However, its linear progression makes it a candidate for early screening programs, and the earlier its detection the higher the chance of recovery. However, one key hurdle for breast cancer screening is the fact that most screening techniques are expensive, time-consuming, and cumbersome, making them impractical for use in several parts of the world. One current trend in breast cancer detection has pointed to a possible solution, the use of salivary breast cancer biomarkers. Saliva is an attractive medium for diagnosis because it is readily available in large quantities, easy to obtain at low cost, and contains all the biomarkers present in blood, albeit in lower quantities. Affinity sensors are devices that detect molecules through their interactions with biological recognition molecules. Their low cost, high sensitivity, and selectivity, as well as rapid detection time make them an attractive alternative to traditional means of detection. In this review article, we discuss the current status of breast cancer diagnosis, its salivary biomarkers, as well as the current trends in the development of affinity sensors for their detection.

Download Full-text

NEWS AND ANNOUNCEMENTS

PEDIATRICS ◽

10.1542/peds.11.3.304 ◽

1953 ◽

Vol 11 (3) ◽

pp. 304-305

Keyword(s):

Medical School ◽

General Hospital ◽

Harvard Medical School ◽

Massachusetts General Hospital ◽

Medical Center ◽

Memorial Hospital

Announcement is made of a 2 Week Postgraduate Course in General Pediatrics, Burnham Memorial Hospital for Children, Massachusetts General Hospital, Boston, under the auspices of the Harvard Medical School, June 8 through June 20, 1953. Clinics providing a broad review of pediatrics emphasizing the new advances will be given daily from 9:00 A.M. to 4:30 P.M. by selected authorities from the Pediatric and Specialty Services of the Massachusetts General Hospital and the Children's Medical Center, Boston.

Download Full-text