New Ultrasound Technique Accurately Distinguishes Breast Cysts From Solid Masses: New Study Shows
USA: A novel ultrasound-based technique may significantly improve the ability of clinicians to differentiate fluid-filled breast cysts from solid masses, potentially reducing unnecessary biopsies and patient anxiety.
The findings are from a study published on December 17 in Radiology Advances by Dr. Muyinatu Bell of Johns Hopkins University, Baltimore, and her colleagues.
Conventional breast ultrasound using B-mode imaging often struggles to clearly distinguish complicated cysts from solid hypoechoic masses because of their similar appearance. This limitation can lead to false-positive assessments and potentially result in additional invasive procedures. To address this challenge, the researchers evaluated an objective, signal-processing–driven metric known as the generalized contrast-to-noise ratio (gCNR), applied to short-lag spatial coherence (SLSC) ultrasound images.
Unlike traditional imaging, the coherence-based approach focuses on how similar ultrasound signals are to those surrounding them. This similarity is quantified into numerical values, allowing each breast mass to be characterized more objectively. The investigators hypothesized that this method would enhance radiologists’ ability to identify suspicious lesions and distinguish fluid from solid content more reliably.
The analysis was conducted as part of a secondary evaluation of the Advanced Ultrasound Signal Processing of Suspicious Breast Images (AUSPICIOUS) observational study led by Johns Hopkins University. Women enrolled between March 2018 and October 2023 were scheduled for ultrasound-guided procedures or follow-up of at least one breast mass. Raw ultrasound data were collected using a research scanner and later processed with custom software developed by the team.
For the final assessment, the researchers analyzed 145 breast masses from 115 women with a mean age of 52 years. Among these, 16 were complicated cysts, and 96 were solid masses. Six radiologists independently defined regions of interest on the SLSC images, and gCNR values were calculated for each lesion. These results were compared with the same radiologists’ interpretations using standard B-mode ultrasound, where masses were classified as solid, fluid, mixed, or uncertain.
The following were the key findings of the study:
- Applying gCNR to SLSC ultrasound images demonstrated excellent diagnostic performance, with a mean AUC of 0.96 for differentiating complicated cysts from solid breast masses.
- Conventional B-mode ultrasound interpretations showed markedly lower accuracy, with a mean AUC of 0.67, highlighting a statistically significant performance gap between the two approaches.
- Complicated cysts that were initially classified as BI-RADS 3 or 4 on standard imaging were more accurately recognized as fluid-filled lesions using the gCNR-SLSC method.
- Use of the objective gCNR metric improved inter-reader agreement, increasing consistency among radiologists from fair with B-mode imaging to moderate with gCNR applied to SLSC images.
- The improved agreement indicates reduced variability between readers, alongside enhanced overall diagnostic reliability.
"Overall, the study demonstrates that incorporating gCNR into SLSC ultrasound imaging offers a more reliable and objective way to differentiate complicated cysts from solid breast masses. If validated further, this approach could support more confident clinical decision-making and help limit unnecessary interventions in breast imaging practice," the authors concluded.
Reference:
Sharma, A., Oluyemi, E. T., Tripathi, M., Ambinder, E. B., Mullen, L. A., Panigrahi, B., Rossi, J., Venkatayogi, N., Myers, K. S., & Lediju Bell, M. A. (2025). Generalized contrast-to-noise ratio applied to short-lag spatial coherence ultrasound differentiates breast cysts from solid masses. Radiology Advances, 2(6). https://doi.org/10.1093/radadv/umaf037
Next Story