Breaking News
March 23, 2019 - The Current issue of “The view from here” is concerned with Cancer
March 23, 2019 - Mouse model validates how ‘good’ and ‘bad’ bacteria affect acne
March 23, 2019 - Individual amygdala neurons respond to touch, imagery and sounds
March 23, 2019 - Combination of two topical creams can prevent cancer
March 23, 2019 - Study suggests depression screening when assessing African-Americans for schizophrenia
March 23, 2019 - New electronic support system for choosing drug treatment based on patient’s genotype
March 23, 2019 - First-of-its-kind study provides pregnancy statistics of imprisoned U.S. women
March 23, 2019 - Marinus Pharmaceuticals Initiates Phase 3 Study in Children with PCDH19-Related Epilepsy
March 23, 2019 - Laparoscopy: MedlinePlus Lab Test Information
March 23, 2019 - Shellfish allergies: can they be treated?
March 23, 2019 - Toilet seat heart monitoring system
March 23, 2019 - Researchers identify way to improve common treatment for PTSD
March 23, 2019 - High potency cannabis use linked to psychosis finds study
March 23, 2019 - Evoke Pharma Submits Response to FDA Review Letter for Gimoti NDA
March 23, 2019 - Tracking HIV’s ever-evolving genome in effort to prioritize public health resources
March 23, 2019 - Scientists grow most sophisticated brain organoid to date
March 23, 2019 - ADHD drug raising risk of psychosis
March 22, 2019 - FDA approves brexanolone, first drug developed to treat postpartum depression
March 22, 2019 - Gruesome cat and dog experiments by the USDA exposed
March 22, 2019 - Ball pits used in children’s physical therapy may contribute to germ transmission
March 22, 2019 - Long-term use of inexpensive weight-loss drug may be safe and effective
March 22, 2019 - FDA Approves Sunosi (solriamfetol) for Excessive Daytime Sleepiness Associated with Narcolepsy or Obstructive Sleep Apnea
March 22, 2019 - Anti-Müllerian Hormone Test: MedlinePlus Lab Test Information
March 22, 2019 - Finding the right exercise, diet aids for HIV patients
March 22, 2019 - Health Plans For State Employees Use Medicare’s Hammer On Hospital Bills
March 22, 2019 - Researchers develop new tool for imaging large groups of neurons in living animals
March 22, 2019 - Certain bacteria and immune factors in vagina may cause or protect against preterm birth
March 22, 2019 - Novel breath test could pave new way to non-invasively measure gut health
March 22, 2019 - Pharmaceutical and personal care products may result in new contaminants in waterways
March 22, 2019 - ACC: Catheter Ablation Does Not Cut Mortality, Stroke in A-Fib
March 22, 2019 - Wiedemann-Rautenstrauch syndrome – Genetics Home Reference
March 22, 2019 - Health insurance is not assurance of healthcare
March 22, 2019 - Supporting “curiosity-driven research” at the Discovery Innovation Awards
March 22, 2019 - Must-Reads Of The Week (Some Flying Below The Radar)
March 22, 2019 - Newly engineered nanoscale protein micelles can be tracked by MRI
March 22, 2019 - New model more effective in predicting risk of opioid overdose than traditional models
March 22, 2019 - Mayo Clinic study identifies potential new drug therapy for liver diseases
March 22, 2019 - Pitt engineers win $550,000 NSF CAREER award to develop new intervention for people with ASD
March 22, 2019 - Early discharge does not increase readmission risk for patients after lung surgery
March 22, 2019 - Creating diverse pool of trained scientists to address Alzheimer’s research needs
March 22, 2019 - Surprising discovery offers clues to limit graft-vs.-host disease
March 22, 2019 - Study shows ACA’s positive impact on healthcare affordability and access for women
March 22, 2019 - Study provides new pathway for controlling inflammation
March 22, 2019 - New combination treatment shows promise for common brain tumor in children
March 22, 2019 - Virginia Tech Helmet Lab releases first youth-specific football helmet ratings
March 22, 2019 - New algae-based treatment could reduce need for limb amputation
March 22, 2019 - Stroke risk reduces in both black and white older Medicare beneficiaries, study reports
March 22, 2019 - City of Hope exhibits current studies and data on cancer therapies at AACR
March 22, 2019 - New study identifies CD40 molecule as key entry point for dangerous bacteria
March 22, 2019 - Health Tip: Six Steps to a Healthier Life
March 22, 2019 - even a little activity helps you live longer
March 22, 2019 - Many individuals recovering from addiction continue to suffer from chronic physical disease
March 22, 2019 - New drugs on PBS for Parkinson’s, MND and Cutaneous T cell lymphoma
March 22, 2019 - Saving energy also saves lives, UW-Madison study says
March 22, 2019 - Former inmates who receive social support have better mental health, study finds
March 22, 2019 - Nanofibrous membrane could enhance periodontal tissue regeneration
March 22, 2019 - Anti-vaxxer Italian leader down with chickenpox
March 22, 2019 - Servier collaborates with Harvard researchers to fight metabolic diseases
March 22, 2019 - National Eating Disorders Association
March 22, 2019 - Pumping up red blood cell production
March 22, 2019 - Excessive phosphate fertilizer may hurt plants by altering microbial composition in soil
March 22, 2019 - Medical marijuana laws could be improving older Americans’ health, study suggests
March 22, 2019 - Study indicates the benefits of stopping aspirin in heart attack patients
March 22, 2019 - Deep brain stimulation offers significant relief for patients with treatment-resistant depression
March 22, 2019 - Mental health problems in young adults on the rise
March 22, 2019 - Innovative membrane offers a viable solution for periodontitis
March 22, 2019 - The FDA Grants Rare Pediatric Disease Designation to Odiparcil for the Treatment of MPS VI
March 22, 2019 - insulin therapy: MedlinePlus Medical Encyclopedia
March 22, 2019 - Guidelines on the use of genetic testing in psychiatry
March 22, 2019 - Aspiring Doctors Seek Advanced Training In Addiction Medicine
March 22, 2019 - A change in focus could enable the development of new treatments for Alzheimer’s disease
March 22, 2019 - A new way to visualize the immune cell “landscape” of bowel cancer tumors
March 22, 2019 - Understanding maintenance of quiescent stem cells in chronic myelogenous leukemia
March 22, 2019 - Ludwig scientists to share advances in cancer research at AACR Annual Meeting 2019
March 22, 2019 - Less invasive valve replacement can be safe and effective alternative for healthier patients
March 22, 2019 - Aphasia research reveals new, complex interactions between thought and language
March 22, 2019 - Artificial neural networks can predict how different areas in the brain respond to words
March 22, 2019 - Age-related changes to gut microbiome have adverse impact on vascular health, study shows
March 22, 2019 - Study provides new insight into blood cell and immune cell production
March 22, 2019 - Isolated seniors chat online to prevent cognitive decline
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