Breaking News
January 19, 2019 - FDA Approves Ontruzant (trastuzumab-dttb), a Biosimilar to Herceptin
January 19, 2019 - Tobacco use linked with higher use of opioids and sedatives
January 19, 2019 - Study delves deeper into developmental dyslexia
January 19, 2019 - Anti-vaccination movement one of the top health threats in 2019 says WHO
January 19, 2019 - Newly developed risk score more effective at identifying type 1 diabetes
January 19, 2019 - Highly effective protocol to prepare cannabis samples for THC/CBD analysis
January 19, 2019 - Prinston Pharmaceutical Inc. Issues Voluntary Nationwide Recall of Irbesartan and Irbesartan HCTZ Tablets Due to Detection of a Trace Amount of Unexpected Impurity, N-Nitrosodiethylamine (NDEA) in the Products
January 19, 2019 - How does solid stress from brain tumors cause neuronal loss, neurologic dysfunction?
January 19, 2019 - $14.7 million partnership to supercharge vaccine development
January 19, 2019 - Ian Fotheringham receives Charles Tennant Memorial Lecture award
January 19, 2019 - Brain vital signs detect neurophysiological impairments in players with concussions
January 19, 2019 - Lack of job and poor housing conditions increased likelihood of people attending A&E
January 19, 2019 - Novel targeted drug delivery system improves conventional cancer treatments
January 19, 2019 - Rutgers study finds gene responsible for spread of prostate cancer
January 19, 2019 - Complications Higher Than Expected for Invasive Lung Tests
January 19, 2019 - 3-D printed implant promotes nerve cell growth to treat spinal cord injury
January 19, 2019 - Automated texts lead to improved outcomes after total knee or hip replacement surgery
January 19, 2019 - Poor cardiorespiratory fitness could increase risk of future heart attack, finds new study
January 19, 2019 - Drinking soft drinks while exercising in hot weather may increase risk of kidney disease
January 19, 2019 - Formlabs 3D prints anatomical models
January 19, 2019 - Heart-Healthy Living Also Wards Off Type 2 Diabetes
January 19, 2019 - Teaching Kids to Be Smart About Social Media (for Parents)
January 19, 2019 - Metabolite produced by gut microbiota from pomegranates reduces inflammatory bowel disease
January 19, 2019 - Researchers examine how spray from showers and toilets expose us to disease causing bacteria
January 19, 2019 - Behavioral experiments confirm that additional neurons improve brain function
January 19, 2019 - New study compares performance of real-time infectious disease forecasting models
January 19, 2019 - Obesity can be risk factor for developing renal cell carcinoma, confirms study
January 19, 2019 - New regulation designs on cigarette packs direct smokers’ attention to health warnings
January 19, 2019 - QIAGEN receives first companion diagnostic approval in Japan
January 19, 2019 - Study explores role of Dunning-Kruger effect in anti-vaccine attitudes
January 19, 2019 - Newly identified subset of immune cells may be key to fighting chronic inflammation
January 19, 2019 - New immune response regulators discovered
January 18, 2019 - Poor blood oxygenation during sleep predicts chance of heart-related death
January 18, 2019 - First international consensus on the diagnosis and management of fibromuscular dysplasia
January 18, 2019 - Rapid resistance gene sequencing technology can hasten identification of antibiotic-resistant bacteria
January 18, 2019 - Researchers develop artificial enzymatic pathway for synthesizing isoprenoids in E. coli
January 18, 2019 - Scientists advise caution in immunotherapy research
January 18, 2019 - How children across the world develop language
January 18, 2019 - Columbia Medical Student Receives McDonogh Scholarship
January 18, 2019 - Secretive ‘Rebate Trap’ Keeps Generic Drugs For Diabetes And Other Ills Out Of Reach
January 18, 2019 - Plant based diet could be the best option for the planet says commission
January 18, 2019 - New conservation practice could reduce nitrogen from agricultural drainage, study shows
January 18, 2019 - UIC researchers receive $1.7 million NCI grant to study Southeast Asian fruit
January 18, 2019 - New study determines the fate of DNA derived from genetically modified food
January 18, 2019 - Scientists develop new gene therapy that prevents axon destruction in mice
January 18, 2019 - Study finds critically low HPV vaccination rates among younger adolescents in the U.S.
January 18, 2019 - Brain cells involved in memory play key role in reducing future eating behavior
January 18, 2019 - Risk for Conversion of MS Varies With Different Therapies
January 18, 2019 - Investigational cream may help patients with inflammatory skin disease
January 18, 2019 - Medical school news office receives six writing awards | News Center
January 18, 2019 - County By County, Researchers Link Opioid Deaths To Drugmakers’ Marketing
January 18, 2019 - Research reveals risk for developing more than one mental health disorder
January 18, 2019 - Scientists discover a dramatic pattern of bone growth in female mice
January 18, 2019 - Study finds link between lengthy periods of undisturbed maternal sleep and stillbirths
January 18, 2019 - New nuclear medicine method could improve detection of primary and metastatic melanoma
January 18, 2019 - Combination therapy shows high efficacy in treating people with leishmaniasis and HIV
January 18, 2019 - Health Tip: Don’t Ignore Changes in Skin Color
January 18, 2019 - Dietary Recommendations for Healthy Children
January 18, 2019 - Eliminating the latent reservoir of HIV
January 18, 2019 - Pain From The Government Shutdown Spreads. This Time It’s Food Stamps
January 18, 2019 - Newly discovered regulatory mechanism helps control fat metabolism
January 18, 2019 - New rapid blood tests could speed up TB diagnosis, save the NHS money
January 18, 2019 - Researchers develop intelligent system for ‘tuning’ powered prosthetic knees
January 18, 2019 - Monoclonal antibody pembrolizumab prolongs survival in patients with squamous cell carcinoma
January 18, 2019 - Maintaining an active lifestyle in older age could prevent dementia
January 18, 2019 - New research detects mosquito known to transmit malaria for the first time in Ethiopia
January 18, 2019 - Researchers identify new genes linked to development of age-related macular degeneration
January 18, 2019 - Computerized method helps better protect pharma patents
January 18, 2019 - New guidelines to make swallowing safer for people in Australian nursing homes
January 18, 2019 - Lumex Instruments’ RA-915AM monitor installed at Hg treatment plant in Almadén, Spain
January 18, 2019 - ACCC survey finds multiple threats to growth of cancer programs
January 18, 2019 - Meeting the challenge of engaging men in HIV prevention and treatment
January 18, 2019 - Furloughed Feds’ Health Coverage Intact, But Shutdown Still Complicates Things
January 18, 2019 - Experts discuss various aspects on health risks posed by fumigated containers
January 18, 2019 - Researchers use gene-editing tool CRISPR/Cas9 to limit impact of parasitic diseases
January 18, 2019 - Alpha neurofeedback training could be a means of enhancing learning success
January 18, 2019 - Innovative ‘light’ method demonstrates positive results in fight against malignant tumors
January 18, 2019 - The cytoskeleton of neurons found to play role in Alzheimer’s disease
January 18, 2019 - New resource-based approach to improve HIV care in low- and middle-income countries
January 18, 2019 - Bedfont appoints Dr Jafar Jafari as first member of the Gastrolyzer Medical Advisory Board
New automated model identifies dense breast tissue in mammograms

New automated model identifies dense breast tissue in mammograms

image_pdfDownload PDFimage_print

Researchers from MIT and Massachusetts General Hospital have developed an automated model that assesses dense breast tissue in mammograms — which is an independent risk factor for breast cancer — as reliably as expert radiologists.

This marks the first time a deep-learning model of its kind has successfully been used in a clinic on real patients, according to the researchers. With broad implementation, the researchers hope the model can help bring greater reliability to breast density assessments across the nation.

It’s estimated that more than 40 percent of U.S. women have dense breast tissue, which alone increases the risk of breast cancer. Moreover, dense tissue can mask cancers on the mammogram, making screening more difficult. As a result, 30 U.S. states mandate that women must be notified if their mammograms indicate they have dense breasts.

But breast density assessments rely on subjective human assessment. Due to many factors, results vary — sometimes dramatically — across radiologists. The MIT and MGH researchers trained a deep-learning model on tens of thousands of high-quality digital mammograms to learn to distinguish different types of breast tissue, from fatty to extremely dense, based on expert assessments. Given a new mammogram, the model can then identify a density measurement that closely aligns with expert opinion.

“Breast density is an independent risk factor that drives how we communicate with women about their cancer risk. Our motivation was to create an accurate and consistent tool, that can be shared and used across health care systems,” says second author Adam Yala, a PhD student in MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL).

The other co-authors are first author Constance Lehman, professor of radiology at Harvard Medical School and the director of breast imaging at the MGH; and senior author Regina Barzilay, the Delta Electronics Professor at CSAIL and the Department of Electrical Engineering and Computer Science at MIT.

Mapping density

The model is built on a convolutional neural network (CNN), which is also used for computer vision tasks. The researchers trained and tested their model on a dataset of more than 58,000 randomly selected mammograms from more than 39,000 women screened between 2009 and 2011. For training, they used around 41,000 mammograms and, for testing, about 8,600 mammograms.

Each mammogram in the dataset has a standard Breast Imaging Reporting and Data System (BI-RADS) breast density rating in four categories: fatty, scattered (scattered density), heterogeneous (mostly dense), and dense. In both training and testing mammograms, about 40 percent were assessed as heterogeneous and dense.

During the training process, the model is given random mammograms to analyze. It learns to map the mammogram with expert radiologist density ratings. Dense breasts, for instance, contain glandular and fibrous connective tissue, which appear as compact networks of thick white lines and solid white patches. Fatty tissue networks appear much thinner, with gray area throughout. In testing, the model observes new mammograms and predicts the most likely density category.

Matching assessments

The model was implemented at the breast imaging division at MGH. In a traditional workflow, when a mammogram is taken, it’s sent to a workstation for a radiologist to assess. The researchers’ model is installed in a separate machine that intercepts the scans before it reaches the radiologist, and assigns each mammogram a density rating. When radiologists pull up a scan at their workstations, they’ll see the model’s assigned rating, which they then accept or reject.

“It takes less than a second per image … [and it can be] easily and cheaply scaled throughout hospitals.” Yala says.

On over 10,000 mammograms at MGH from January to May of this year, the model achieved 94 percent agreement among the hospital’s radiologists in a binary test — determining whether breasts were either heterogeneous and dense, or fatty and scattered. Across all four BI-RADS categories, it matched radiologists’ assessments at 90 percent. “MGH is a top breast imaging center with high inter-radiologist agreement, and this high quality dataset enabled us to develop a strong model,” Yala says.

In general testing using the original dataset, the model matched the original human expert interpretations at 77 percent across four BI-RADS categories and, in binary tests, matched the interpretations at 87 percent.

In comparison with traditional prediction models, the researchers used a metric called a kappa score, where 1 indicates that predictions agree every time, and anything lower indicates fewer instances of agreements. Kappa scores for commercially available automatic density-assessment models score a maximum of about 0.6. In the clinical application, the researchers’ model scored 0.85 kappa score and, in testing, scored a 0.67. This means the model makes better predictions than traditional models.

In an additional experiment, the researchers tested the model’s agreement with consensus from five MGH radiologists from 500 random test mammograms. The radiologists assigned breast density to the mammograms without knowledge of the original assessment, or their peers’ or the model’s assessments. In this experiment, the model achieved a kappa score of 0.78 with the radiologist consensus.

Next, the researchers aim to scale the model into other hospitals. “Building on this translational experience, we will explore how to transition machine-learning algorithms developed at MIT into clinic benefiting millions of patients,” Barzilay says. “This is a charter of the new center at MIT — the Abdul Latif Jameel Clinic for Machine Learning in Health at MIT — that was recently launched. And we are excited about new opportunities opened up by this center.”

Source:

http://news.mit.edu/2018/AI-identifies-dense-tissue-breast-cancer-mammograms-1016

Tagged with:

About author

Related Articles