News | Breast Imaging | January 28, 2021

Mammogram-based Breast Cancer Risk Model Could Lead to Better Screening Guidelines

The four standard views of an individual mammogram were fed into Mirai. The image encoder mapped each view to a vector, and the image aggregator combined the four view vectors into a single vector for the mammogram. In this work, we used a single shared ResNet-18 as an image encoder, and a transformer as our image aggregator. The risk factor predictor module predicted all the risk factors used in the Tyrer-Cuzick model, including age, detailed family history, and hormonal factors, from the mammogram vector.

The four standard views of an individual mammogram were fed into Mirai. The image encoder mapped each view to a vector, and the image aggregator combined the four view vectors into a single vector for the mammogram. In this work, we used a single shared ResNet-18 as an image encoder, and a transformer as our image aggregator. The risk factor predictor module predicted all the risk factors used in the Tyrer-Cuzick model, including age, detailed family history, and hormonal factors, from the mammogram vector. The additive hazard layer combined information from both the image aggregator and risk factors (predicted or given) to predict coherent risk assessments across 5 years (Yr).

January 28, 2021 — A new machine learning algorithm based on mammograms can estimate the risk of breast cancer in women more accurately than current risk models, according to a study from Adam Yala and colleagues. The algorithm, which was tested with datasets from three large hospitals located worldwide, could help clinicians design guidelines for breast cancer screening that meet the need for early detection while reducing false-positives, test costs, and other issues associated with overscreening. Mammograms are the most common method to screen for breast cancer, as more than 39 million procedures are performed in the U.S. annually. However, their widespread adoption has not gone without controversy. Critics charge that aggressive screening results in untenable medical costs, higher anxiety in patients, and a substantial rate of false positives. On the other hand, supporters of frequent testing argue it is necessary to detect tumors as early as possible, and the disagreement has led to inconsistent guidelines for when screening should start and how frequent it should be. Yala et al. theorized that improving the accuracy of the risk models that inform guidelines could lead to better recommendations. They designed and trained a new model named Mirai, which integrates data from mammograms to produce consistent breast cancer risk assessments at multiple time points, such as within 1 year or 5 years. When independently tested with data from 106,615 patients from three hospitals, located in the U.S., Sweden, and Taiwan, Mirai identified 41.5% of patients who would develop cancer within 5 years. In contrast, current approaches such as the Tyrer-Cuzick and Hybrid Deep Learning models only identified 22.9% and 36.1% of patients, respectively. Mirai was also effective across various races and ethnicities, supporting its potential to inform screening guidelines for large and diverse populations.

For more information: www.aaas.org

Related Content

Artificial intelligence (AI)-driven healthcare has potential to transform medical decision-making and treatment, but AI algorithms must be thoroughly tested and continuously monitored to avoid unintended consequences to patients. In JAMA Network Open, Regenstrief Institute President Peter Embí, M.D., calls for algorithmovigilance (a term he coined for scientific methods and activities relating to evaluation, monitoring, understanding and prevention of adverse effects of algorithms in healthcare) to address

Artificial intelligence (AI)-driven healthcare has potential to transform medical decision-making and treatment, but AI algorithms must be thoroughly tested and continuously monitored to avoid unintended consequences to patients. In JAMA Network Open, Regenstrief Institute President Peter Embí, M.D., calls for algorithmovigilance (a term he coined for scientific methods and activities relating to evaluation, monitoring, understanding and prevention of adverse effects of algorithms in healthcare) to address inherent biases in healthcare algorithms and their deployment. Image courtesy of Regenstrief Institute

News | Artificial Intelligence | May 14, 2021
May 14, 2021 — Artificial intelligence
Chest X-ray from patient severely ill from COVID-19, showing (in white patches) infected tissue spread across the lungs. Image courtesy of Nature Publishing or npj Digital Medicine

Chest X-ray from patient severely ill from COVID-19, showing (in white patches) infected tissue spread across the lungs. Image courtesy of Nature Publishing or npj Digital Medicine

News | Coronavirus (COVID-19) | May 14, 2021
May 14, 2021 — Trained to see patterns by analyzing thousands of chest...
Herman Oosterwijk

Herman Oosterwijk

News | Artificial Intelligence | May 13, 2021
May 13, 2021 — Laurel Bridge Software, a provider of ...
Medical researchers at Flinders University have established a new link between high body mass index (BMI) and breast cancer survival rates — with clinical data revealing worse outcomes for early breast cancer (EBC) patients and improved survival rates in advanced breast cancer (ABC).

Getty Images

News | Women's Health | May 13, 2021
May 13, 2021 — Medical researchers at Flinders University
 Novarad, healthcare IT and medical imaging company, has recently entered into contract with Waverly Health Center (WHC) of Iowa to provide its award-winning solution for simply and securely sharing medical images and reports – CryptoChart.
News | Enterprise Imaging | May 12, 2021
May 12, 2021 — Novarad, healthcare IT and medical imaging company, has recently entered into contract with...
News | Artificial Intelligence | May 11, 2021
May 11, 2021 — Whiterabbit, an AI technology company focused on early...