Sponsored Content | Case Study | Mammography | March 06, 2018

Southwest Diagnostic Imaging Center Uses Artificial Intelligence to Optimize Breast Tomosynthesis Interpretations

Katherine Hall, M.D.

Katherine Hall, M.D.

Located in a spacious 25,000-square-foot facility on the Presbyterian Hospital of Dallas campus, Southwest Diagnostic Imaging Center is one of the largest freestanding imaging centers in the Dallas-Fort Worth area. Since 1985, the facility has provided thousands of patients from Texas and across the nation with the highest standard of service and a patient-centered approach to care.

 

Leading the Way

While the adoption of breast tomosynthesis is expanding, so is the innovative technology that supports these solutions. Southwest Diagnostic Imaging Center has been paving the way, equipping its radiologists with the most advanced imaging solutions for more than 30 years, and has now added artificial intelligence solution PowerLook Tomo Detection to their women’s breast health program.

To maintain competitive advantage, transitioning from 2-D to 3-D mammography, or digital breast tomosynthesis (DBT), was the imaging center’s next step. However, with nearly 12,000 mammograms conducted each year, finding a cutting-edge solution that ensured both accuracy and efficiency was crucial.

“Previously running a facility that used DBT for many years, I knew first-hand the incredible benefits and challenges radiologists face with the technology,” said Katherine Hall, M.D., radiologist and co-director of mammography, east division, at Southwest Diagnostic Imaging Center. “While I was eager to make the switch from 2-D to DBT to better serve our patients, I also expected some internal resistance, as going from reading just a few images in 2-D to reading hundreds of images per patient with DBT can be both tiresome and time-intensive for radiologists.”

 

Conducting the Search

In May 2016 Southwest Diagnostic Imaging Center conducted an extensive review of all existing tomosynthesis technologies on the market, ultimately deciding to implement GE Healthcare’s DBT solution. In doing so, the facility was also able to take advantage of iCAD’s PowerLook Tomo Detection.

Specifically, PowerLook Tomo Detection reduces the need to scroll through hundreds of images. Rather, it utilizes deep learning to pick up the highest densities and areas of distortion on a mammogram, extracts them and blends them onto a synthetic 2-D image.

“Instead of looking at 80-90 planes or more per projection for those patients with large breasts, PowerLook Tomo Detection leads me directly to the images that are most critical to review,” said Hall. “I am then able to click on that area in the preview and the solution takes me directly to the location needing further evaluation. In fact, any suspicious masses or distortions on the image are so clear that they literally pop out at you.”

 

Reducing Read Time, Improving Workflow

After four months, Southwest Diagnostic Imaging Center’s radiologists are reaping the benefits of PowerLook Tomo Detection. In fact, not only did PowerLook Tomo Detection help Hall move to an all 3-D protocol, she’s also cut down her reading time anywhere from 25-50 percent. “We want to ensure quality care for our patients and with the adoption I feel confident to utilize a 3-D only protocol. This is a benefit to our patients as it is a reduction in radiation dose exposure,”she said.

“As radiologists, when we’re amassed with information and are in a state of image overload, we all need help seeing,” continued Hall. “PowerLook Tomo Detection has been a game changer. It helps me collimate and focus on what I really need to look at and evaluate, which is what I rely on to provide my patients with the best possible care.”

Case study supplied by iCAD.

Related Content

VolparaDensity With Tyrer-Cuzick Model Improves Breast Cancer Risk Stratification
News | Breast Density | May 22, 2019
Research has demonstrated use of Volpara Solutions' VolparaDensity software in combination with the Tyrer-Cuzick Breast...
Partial Breast Irradiation Effective, Convenient Treatment Option for Low-Risk Breast Cancer
News | Radiation Therapy | May 20, 2019
Partial breast irradiation produces similar long-term survival rates and risk for recurrence compared with whole breast...
Managing Architectural Distortion on Mammography Based on MR Enhancement
News | Mammography | May 15, 2019
High negative predictive values (NPV) in mammography architectural distortion (AD) without ultrasonographic (US)...
Screening MRI Detects BI-RADS 3 Breast Cancer in High-risk Patients
News | MRI Breast | May 09, 2019
When appropriate, short-interval follow-up magnetic resonance imaging (MRI) can be used to identify early-stage breast...
Novel Artificial Intelligence Method Predicts Future Risk of Breast Cancer
News | Artificial Intelligence | May 08, 2019
Researchers from two major institutions have developed a new tool with advanced artificial intelligence (AI) methods to...
Georgia Becomes 38th State With Breast Density Inform Law
News | Breast Density | May 06, 2019
Georgia Gov. Brian Kemp signed breast density inform bill HB62 into law May 3, 2019. The law, which becomes effective...
American Society of Breast Surgeons Issues Risk-based Screening Mammography Guidelines
News | Mammography | May 03, 2019
The American Society of Breast Surgeons (ASBrS) has released the first screening mammography guidelines based on a...
Hologic Launches Trident HD Specimen Radiography System
Technology | Breast Biopsy Systems | April 26, 2019
Hologic Inc. announced global commercial availability of the Trident HD specimen radiography system. The next-...
California Protons Cancer Therapy Center Begins Breast Cancer Study Enrollment
News | Proton Therapy | April 25, 2019
California Protons Cancer Therapy Center announced two major efforts representing significant steps forward in breast...
Women With Coronary Artery Wall Thickness at Risk for Heart Disease
News | Cardiac Imaging | April 25, 2019
The thickness of the coronary artery wall as measured by magnetic resonance imaging (MRI) is an independent marker for...