AI's New Vision: DeepHealth's Ultrasound Tech Clears FDA Approval

SOMERVILLE, Mass. – July 30, 2026 – In a significant step toward integrating artificial intelligence into the core of diagnostic medicine, DeepHealth, Inc., the AI-focused subsidiary of imaging giant RadNet, has received FDA 510(k) clearance for its AI-powered breast ultrasound solution. The clearance paves the way for a technology designed to bring unprecedented standardization, accuracy, and efficiency to one of the most critical, yet operator-dependent, tools in breast cancer detection.

This new system, commercially known as DeepHealth Breast Ultrasound, automates the detection and characterization of potential lesions, streamlines reporting, and aims to shorten exam times. For the nearly 40% of women who undergo a breast ultrasound, this innovation represents a potential shift from a highly variable procedure to a more consistent and reliable standard of care, powered by intelligent algorithms.

A New Standard in a Complex Exam

Breast ultrasound has long been an essential, albeit challenging, component of the breast care pathway. It serves as a vital supplemental screening tool, especially for women with dense breast tissue where mammograms can be less effective. However, the quality and interpretation of an ultrasound are heavily reliant on the skill and experience of the individual sonographer performing the exam and the radiologist reading it. This variability can lead to inconsistencies in image acquisition and diagnosis.

DeepHealth's solution, cleared under the technical name See-Mode Augmented Reporting Tool, Breast (SMART-B), is engineered to address this fundamental challenge. The AI platform assists clinicians by automatically localizing suspicious soft-tissue lesions, analyzing their features in line with the established ACR BI-RADS classification system, and generating comprehensive reports. This automation doesn't replace the expert clinician but rather empowers them with a powerful, consistent assistant.

“Breast ultrasound is an essential component of the breast care pathway... It is a highly complex, operator-dependent examination, which can lead to significant variability in image acquisition, interpretation and reporting,” said Dr. Jason McKellop, Medical Director of Women's Imaging for RadNet California. “With DeepHealth's breast ultrasound solution, we can achieve greater standardization of workflows, improving consistency while saving time for patients, sonographers and radiologists.”

The clinical data submitted to the FDA underscores the technology's potential impact. A multi-reader, multi-case study involving 16 U.S. board-certified radiologists found that the AI demonstrated greater than 98% accuracy in localizing breast lesions. More critically, it improved the sensitivity for breast cancer detection by 8%—a meaningful increase that could translate to earlier diagnoses and better patient outcomes. The study also quantified the efficiency gains, showing a remarkable 37% reduction in interpretation time for radiologists, freeing them to focus on complex cases and patient care.

Expanding the Circle of Care for Patients

While the technical and workflow enhancements are significant for providers, the ultimate promise of this innovation lies in its potential to improve patient care, particularly for those in higher-risk categories. The national conversation around breast health has increasingly focused on the limitations of a one-size-fits-all screening approach. For millions of women with dense breasts, a standard mammogram may not be enough, making supplemental imaging like ultrasound a necessity.

DeepHealth's platform is designed to make this supplemental screening more robust and reliable. By standardizing the process, the AI helps ensure that every patient receives the same high level of scrutiny, regardless of where the exam is performed or who performs it. This consistency is a cornerstone of equitable healthcare.

“No single imaging pathway addresses every woman’s needs. With the addition of Breast Ultrasound, we are proud to support women across a broader range of screening and diagnostic pathways, including those with dense breasts and others who may require supplemental imaging,” said Niccolò Stefani, M.D., Business and Product Leader for Clinical AI at DeepHealth. “Bringing together AI-powered capabilities across mammography and ultrasound helps clinicians respond to different imaging needs and deliver more comprehensive, personalized breast care.”

This integration is key. The new ultrasound AI joins DeepHealth’s existing suite of tools for mammography, which includes AI for cancer detection, density assessment, and risk prediction. Together, they form one of the industry's most comprehensive AI-driven platforms for breast health, enabling a more holistic and personalized diagnostic journey for each patient.

The Strategic Blueprint for AI-Driven Diagnostics

The FDA clearance is not just a clinical milestone; it's a major strategic move for DeepHealth and its parent company, RadNet, a leading national provider of diagnostic imaging services. RadNet plans an aggressive rollout, intending to implement the solution across its vast network of imaging centers by the end of the year. This move will immediately bring the technology to scale, potentially impacting the more than 700,000 breast ultrasound studies performed annually within its network.

This rapid deployment is facilitated by RadNet's existing infrastructure and its overarching vision for an AI-integrated future, managed through its cloud-native DeepHealth OS. This operating system is designed to unify data and AI tools across the entire diagnostic workflow, making the adoption of new technologies like the breast ultrasound solution seamless.

Financially, the technology is eligible for reimbursement under an existing Category III CPT code for quantitative ultrasound tissue characterization. While these temporary codes are designed for emerging technologies and reimbursement can vary by payer, RadNet’s scale provides a powerful platform for demonstrating the technology's value and paving the way for more stable, widespread reimbursement in the future. This move solidifies RadNet's position as a leader not just in providing imaging services, but in shaping the technological future of the industry.

Navigating the Path to Widespread Adoption

Despite the clear benefits and a strong strategic push from a major industry player, the path to universal adoption for technologies like DeepHealth Breast Ultrasound involves navigating significant institutional hurdles. For imaging centers outside of large, vertically integrated networks like RadNet, the primary challenges will be cost and reimbursement.

The uncertainty surrounding Category III CPT codes can make the initial investment in new AI software a difficult decision for smaller providers. Without guaranteed payment from insurers, the return on investment remains a calculated risk. Furthermore, integrating new software into legacy IT systems and training staff to trust and effectively utilize AI-powered insights requires a dedicated commitment of time and resources.

However, the momentum behind AI in diagnostics is undeniable. The demonstrated improvements in accuracy, efficiency, and standardization create a compelling case for change. As large networks like RadNet prove the clinical and operational value of these platforms at scale, they build a powerful repository of data that will likely influence payer policies and encourage broader adoption. The efficiency gains, such as a 37% reduction in radiologist reading time, directly address the growing pressures of staff shortages and physician burnout, making AI less of a luxury and more of a necessity for a sustainable healthcare system.