GE HealthCare says its Pristina Recon DL technology uses two deep learning models working in sequence for sharper images with fewer artifacts, enhanced details, and greater consistency. [Photo courtesy of GE HealthCare]
GE HealthCare has won FDA Premarket Authorization (PMA) for its Pristina Recon DL technology for 3D mammography image reconstruction.
GE HealthCare says it’s the first mammography technology that combines deep learning and iterative reconstruction to “provide outstanding digital breast tomosynthesis (DBT) image quality without compromising on patient dose.”
It’s the latest example of how GE Healthcare — the world’s sixth-largest medical device company according to the 2025 Medical Design & Outsourcing Medtech Big 100 and a leading developer of artificial intelligence for medtech — is using AI.
MDO Min-Vasive Medtech: Download our free 72-page special report with interviews from minimally invasive device development experts at major OEMs and groundbreaking startups
The technology builds off the Pristina Via system, using two deep learning algorthms and Nvidia RTX accelerated computing technology for sharper images with fewer artifacts, enhanced details, and greater consistency to help health care providers make more confident breast cancer diagnoses, GE HealthCare said
“Pristina Recon DL was born out of a deep commitment to our customers – listening closely to their feedback and working hand-in-hand with radiologists to enhance image quality and clarity,” GE HealthCare Women’s Health and X-ray President and CEO Jyoti Gupta said in a news release. “By applying advanced deep learning technologies, we’re shaping the future of breast imaging — one defined by uncompromised image quality, faster workflows, and greater confidence in early cancer detection.”
GE HealthCare says its Pristina Via with Recon DL system features inclusive design that “ensures accessibility for all patients, reduces anxiety, and improves comfort with patient-assisted compression.” [Photo courtesy of GE HealthCare]
GE HealthCare enhanced image quality with an update to the DBT reconstruction algorithm, “using deep learning algorithms to correct sparse view artifacts, compensate for patient motion artifacts and enhance the visibility of soft tissue lesions within the 2D synthetic image,” according to FDA records.
The two deep learning models “work in sequence to enable separation of meaningful signal from noise,” GE HealthCare said.
The first deep learning model reconstructs high fidelity 3D volumes with greater purity to minimize artifacts and perceived noise, the device developer said, while the second model enhances the visualization of clinically relevant information in the DL synthesized 2D view.
GE HealthCare said it worked with academic institutions and high-volume outpatient imaging centers to develop and validate the algorithms at each stage of the development process.
The new technology can better detect microcalcifications and masses and has the lowest 3D screening dose among major systems on the market, GE HealthCare said. The system also features zero-click acquisition, personalized protocols to simplify and accelerate exam workflows, and inclusive design that “ensures accessibility for all patients, reduces anxiety, and improves comfort with patient-assisted compression.”
“With Pristina Via with Recon DL, we’re setting a new benchmark in breast imaging — delivering sharper, clearer, and more consistent images that empower radiologists with more confidence,” GE HealthCare Mammography VP and GM Pooja Pathak said. “As an upgradable feature on the Pristina Via platform, we are excited to now offer customers uncompromised image quality combined with fast, accurate workflows.”
Related: GE HealthCare’s chief AI officer offers tips and advice for working with artificial intelligence
link