AI-Powered Decision Support in Breast Diagnostics on the GE Healthcare LOGIQ E10 Ultrasound System

We take a look at how GE’s latest Artificial Intelligence technology is helping to support breast imaging diagnostics.

Artificial intelligence (AI) is driving transformation in a number of fields, from manufacturing to communications. Some of the most exciting applications are in healthcare.

With AI, clinicians now have a powerful tool to harness the data collection power of computers to advance their ability to identify, diagnose, and treat disease. In ultrasound, for example, AI is being used to analyze the large, complex data sets generated in imaging. Its ability to rapidly identify recurring patterns in the data can be invaluable in helping doctors determine the location, classification and severity of conditions. Through AI, ultrasound devices are evolving into highly capable digital assistants for clinicians.

Consider the potential of AI-powered decision support in ultrasound breast diagnostics. Detecting and characterizing breast disease, especially in women with dense tissue, can be challenging. While the Breast Imaging Reporting and Data System (BI-RADS) has helped standardize the classification of breast lesions in ultrasound, clinicians still interpret up to one in three cases differently.[1]

To help address this variability, GE Healthcare designed the LOGIQ™ E10 ultrasound system with an available AI-powered decision-support tool for breast diagnostics, in partnership with Koios Medical. Known as Breast Assistant, powered by Koios DS™, this tool provides an AI-based quantitative risk assessment to augment the clinician’s decision-making.

The GE Healthcare LOGIQ E10 ultrasound system in action

How does it work? 

During an ultrasound exam, the technician spots a suspicious area in the breast image. After contouring the lesion, the technician clicks the Breast Assistant button on the console. The AI tool automatically analyzes the lesion, generating a BI-RADS-aligned category and risk assessment within two seconds or less. The color-coded scale aligns to the likelihood of malignancy—in essence helping to “dial in” the BI-RADS assessment, to support the clinician’s decision-making.

What is the value of this supplementary information? 

Studies suggest it helps clinicians increase the accuracy of their diagnostic decisions and order fewer biopsies that result in benign findings. Breast Assistant uses machine learning and a proprietary algorithm to recognize subtle patterns in breast tissue, based on data from more than 400,000 images. In research published in the Journal of Digital Imaging and Signal Processing in Medicine and Biology Symposium[1],[2], this technology improved accuracy in breast cancer diagnosis for radiologists across all levels of experience:

  • Sensitivity increased from 92-97% to 97-98%
  • Specificity increased from 38-46% to 45-52%
  • Benign biopsy rates were reduced 34-55% without a reduction in sensitivity

Another study found a cancer identification rate of 100% with a 69% reduction in benign biopsies.[3] 

AI-enabled ultrasound devices will never replace the intelligence, skills, and experience of diagnosticians. But they can assume key data analysis tasks and perform them at a speed and scale not possible for humans. As such, tools like Breast Assistant, powered by Koios DS can act as diagnostic sentinels, alerting clinicians to subtle yet meaningful variations in image data to support accurate, timely patient care decisions.

[1] Koios Medical internal data.

[2] Barinov, L., Jairaj, A., Becker, M. et al. Impact of Data Presentation on Physician Performance Utilizing Artificial Intelligence-Based Computer-Aided Diagnosis and Decision Support Systems. J Digit Imaging (2018). https://doi.org/10.1007/s10278-018-0132-5.

[3] Barinov L, Jairaj A, Paster L, Hulbert W, Mammone R, Podilchuk C: Decision quality support in diagnostic breast ultrasound through artificial intelligence. IEEE Signal Processing in Medicine and Biology Symposium (SPMB)., 2016.

[4] Love SM, Berg WA, Podilchuk C, Hovanessian-Larsen LJ, Dauphine C, Jairaj A, Barinov L, Hulbert W, Cen S, Eshraghi L, Mammone R. Automated, low-cost palpable breast lump triage for economically-developing countries [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr PD3-01.

About Koios Medical

Koios Medical develops medical software to assist physicians interpreting ultrasound images and applies deep learning methods to the process of reaching an accurate diagnosis. The Koios DS platform uses advanced algorithms to assist in the early detection of disease while reducing recommendations for biopsy of benign tissue. Patented technology saves physicians time, helps improve patient outcomes, and reduces healthcare costs. To learn more, visit koiosmedical.com.

LOGIQ is a trademark of the General Electric Company.  Koios DS is a trademark of Koios Medical.

Be the first to comment

Leave a Reply

Your email address will not be published.


*