GE HealthCare & NVIDIA Collaborate to Enhance Ultrasound Imaging with AI Technology

GE HealthCare aims to integrate foundational AI technologies into medical imaging to expedite AI application development for clinicians and patients.

GE HealthCare and NVIDIA have joined forces to unveil SonoSAMTrack, an AI-powered research model aimed at revolutionizing ultrasound image segmentation, ultimately leading to improved efficiency and accuracy in patient care.

SonoSAMTrack, developed by GE HealthCare, leverages NVIDIA's accelerated computing and AI technology stack to precisely segment anatomies, lesions, and other critical areas within ultrasound images.

The model is part of a broader effort by GE HealthCare to integrate foundational AI technologies into medical imaging, aiming to speed up AI application development and ultimately benefit clinicians and patients.

GE HealthCare & NVIDIA's Plans for Foundation Models

GE HealthCare emphasizes that integrating AI into healthcare typically involves retraining models to suit diverse patient populations and hospital environments. This retraining process can result in increased expenses, complexity, and the necessity for specialized staff.

On the other hand, foundation models function as human-in-the-loop AI systems, according to GE HealthCare.

These models enable quick adjustments to different diseases, aiding in screening, early detection, monitoring progression, and identifying non-invasive biomarkers with minimal training needs.

Reportedly, this advanced AI model has been put through rigorous testing across seven different ultrasound datasets and has claimed to showcase its versatility in handling various anatomical structures and pathologies.

SonoSAMTrack has proven its effectiveness across seven different ultrasound datasets, showcasing its versatility in handling various anatomies and pathologies, including adult hearts, fetal heads, breast lesions, and musculoskeletal issues.

As per the companies, SonoSAMTrack's performance has surpassed alternative methods by a considerable margin.

Notably, it has demonstrated remarkable speed and efficiency, requiring as few as 2-6 clicks for precise segmentation.

This streamlined process promises to reduce the need for extensive user input, enhancing workflow efficiency for healthcare professionals.

A New Approach to AI Integration

Integrating AI into healthcare has traditionally involved the laborious process of retraining models to suit specific patient groups and hospital settings.

However, foundation models like SonoSAMTrack claim to offer a more agile approach, enabling swift adaptation to a wide range of medical conditions without the need for extensive training.

This flexibility holds the potential to revolutionize medical imaging practices, making AI-powered diagnostics more accessible and efficient.

Sharing thoughts, David Niewolny, director, business development for healthcare and medical, NVIDIA, said, "Combining Nvidia's accelerated computing and AI technology stack with GE HealthCare's medical imaging expertise will help enhance patient care by making ultrasound diagnostics quicker and more accurate. This collaboration underscores the importance of using AI for life-saving advancements and setting new standards in healthcare."

It's essential to note that SonoSAMTrack is still in the development phase and not currently available for commercial sale.

The technology has yet to receive clearance or approval from regulatory bodies such as the U.S. FDA. The information provided is based on a press release statement and reflects ongoing research and development efforts within GE HealthCare.

NVIDIA, a prominent figure in AI computing, collaborated with several medtech firms to integrate AI into their technologies. Johnson & Johnson MedTech recently unveiled its intention to expedite and expand AI implementation in surgery through a partnership with NVIDIA.

Last year, Asensus Surgical joined forces with NVIDIA to provide innovative clinical insights to surgeons using surgical robotics.

Additionally, Medtronic teamed up with NVIDIA to enhance the capabilities of the GI Genius intelligent endoscopy module by introducing an AI Access platform.